The Collapsing Leviathan

I was seriously depressed for the last week, by noticeably more than my baseline amount for the new pandemic-ravaged world. The depression seems to have been triggered by two pieces of news:

  1. The US Food and Drug Administration—yes, the same FDA whose failure to approve covid tests in February infamously set the stage for the deaths of 100,000 Americans—has now also banned the Gates Foundation’s program for at-home covid testing. This, it seems to me, is not the sort of thing that could happen in a still-functioning society, one where people valued their own and their neighbors’ physical survival, and viewed rules and regulations as merely instruments to that end. It’s the sort of thing that one imagines in the waning years of a doomed empire, when no one pretends anymore that they can fix or improve the Leviathan; they’re all just scurrying to flee the Leviathan as it collapses with a thud. More broadly, I still don’t think that the depth of America’s humiliation and downfall has sunk in to most Americans. For me, it starts and ends with a single observation: where fifty years ago we landed humans on the moon, today we can no longer make or distribute paper masks, even when hundreds of thousands of lives depend on it. Look, there are many countries, like Taiwan and New Zealand, that managed to protect both their economies and their vulnerable citizens’ lives, by crushing the virus early. Then there are countries that waited, until they faced an excruciating choice between the two. But here in the US, we’ve somehow achieved the worst of both worlds—triggering a second Great Depression while also utterly failing to control the virus. Can we abandon the charade of treating this as a legible “policy choice,” to be debated in earnest thinkpieces? To me, it just feels like the death-spasm of a collapsing Leviathan.
  2. Something that, at first glance, might seem trivial by comparison, but isn’t: the University of California system—ignoring the advice of its own Academic Senate, and at the apparent insistence of its chancellor Janet Napolitano—will now permanently end the use of the SAT and ACT in undergraduate admissions. This is widely expected, probably correctly, to trigger a chain reaction, whereby one US university after the next will abandon standardized tests. As a result, admissions to the top US universities—and hence, most chances for social advancement in the US—will henceforth be based entirely on shifting and nebulous criteria that rich, well-connected kids and their parents spend most of their lives figuring out, rather than merely mostly based on such criteria. The last side door for smart noncomformist kids is now being slammed shut. From now on, in the US, the only paths to success that clearly delineate their rules will be sports, gambling, reality TV, and the like. In case it matters to anyone reading this, I feel certain that a 15-year-old me wouldn’t stand a chance in the emerging regime—any more than nerdy Jewish kids did in the USSR of the 1970s, or the US of the 1920s. (As I’ve previously recounted on this blog, the US’s “holistic” college admissions system, with its baffling-to-foreigners emphasis on “character,” “leadership,” “well-roundedness,” etc. rather than test scores, originated in a successful push a century ago by the presidents of Harvard, Princeton, and Yale to keep Jewish enrollments down. Today the system fulfills precisely the same function, except against Asian-Americans rather than Jews.) Ironically but predictably, the death of the SAT—i.e., of one of the most fearsome weapons against entrenched wealth and power ever devised—is being celebrated by the self-described champions of the underdog. I have one question for those champions: do you not understand what your system will actually do to society’s underdogs? Or do you understand perfectly well, and approve?

To put it bluntly—since events like these leave no room for euphemism—a hundred thousand Americans are now dead from covid, and hundreds of thousands more are poised to die, because smart people are no longer in charge. And the death of the SAT will help ensure that smart people will never be back in charge. Obama might be remembered by history as America’s last smart-person-in-charge, its last competent technocrat—but one man couldn’t stop a tidal wave of stupid.

I know from experience what many will readers will say to all this: “instead of wallowing in gloom, Scott, why don’t you just make falsifiable predictions about the bad outcomes you expect from these developments, and then score yourself later?”

So here’s the thing about that.

Shortly after Trump was elected, I changed this blog’s background to black, as a small way to mourn the United States that I’d grown up thinking that I lived in, the one that had at least some ideals. Today, with four years of hindsight, my thinking then feels overly optimistic: why plain black? Why not, like, images of rotting corpses in a pit?

And yet, were I foolish enough to register predictions in 2016, I would’ve said that within one year, Trump’s staggering incompetence would surely cause some catastrophe or other to grip the country—a really obvious one, with mass death and even Trump’s beloved stock market cratering.

And then after a year, commenters would ridicule me, because none of that had happened. After two years, they’d ridicule me again because it still hadn’t happened, and after three years they’d ridicule me a third time.

Now it’s happened.

America, we now know, is like the cartoon character who runs off a cliff: it dangled in midair for three years, defying physics, before it finally looked down.

Look, I’m a theoretical computer scientist. By training, I deal in asymptotics, not in constant factors. I don’t often make predictions with deadlines; when I do, I often regret it. It’s a good thing that I became an academic rather than an investor! For I’ve learned that the only “oracular power” I have is to make statements like:

My eyes, my brain, and the pit of my stomach are all blaring at me that the asymptotics of this situation just took a sharp turn for the worse. Sure, for an unknown length of time, noise and constant factors could mask the effects. But eventually, either (1) society will need to reverse what it just did, or else (2) terrible effects will spring from it, or else (3) the entire universe no longer makes sense.

When I’ve felt this way in the past, option (3) rarely turned out to be the right answer.

So, what can anyone say that will make me less depressed? Thanks in advance!

Update (May 30): Woohoo!! Avoiding yet another tragedy, after years of setbacks and struggles, it looks like today the US has finally launched humans into orbit, thereby recapitulating a technological achievement from 1961 that the US had already vastly surpassed by 1969. I hereby retract the pessimism of this post.

184 Responses to “The Collapsing Leviathan”

  1. Anon Says:

    Why do you automatically assume that removing the SAT is a bad idea?

    Just like ETS, College Board is a predatory organization, charging ridiculous amounts of money for tests that fail to accurately measure one’s competence.

    I remember that when I was applying to US schools I had to make some painful financial sacrifices to pay for the SAT’s. Then there was also the issue that they would not organize them near my hometown, so I had to take a 6 hour long train ride to the nearest testing site, pay again for a place to spend the night, take the test, then go back home. When this was over, I had to send the scores to several schools, and for each of these, once again, I had to pay.

    What was the outcome? I went to an amazing school, but the time spent preparing for these tests convinced me that chances of scoring high are reserved to those privileged kids (myself included) who can afford to buy the SAT prep books (or have somewhere to borrow them from), and get to spend a few months tediously mastering ways to beat the test. You can not really hope to get a scholarship without an excellent score.

    And yes, you can beat the test once you learn the tricks. Write an essay that respects a certain structure, and appropriately uses connectors. It doesn’t matter if the arguments you use are true or not, they just have to sound convincing enough. Learn a list of words that are likely to show up in the reading section. If you’re not skilled in math, just get used to the types of questions that appear in practice tests. That’s it.

    In the end, those who benefit from these tests are still the privileged. Others just give up.

  2. Peter Morgan Says:

    Quantum Mechanics and Classical Mechanics can, like, be unified. Koopman’ed, one could say. It’s time for a different cartoon QM. With the slowly gathering realization that quantum woo-woo was no longer under the table to justify anything whatsoever, mid-21st century science and society took a step back from having no towel. No need to be Marvin if one can find a towel.
    Wishful thinking apparently works for me, some of the time. Be well, Scott.

  3. Pku31 Says:

    Possibly optimistic: From the outside view, we can compare to the fall of Rome – Rome had very smart people sure it was in terminal decline for most of its existence, but they were usually wrong – and even when they were right about it being in decline, it was more likely to resurge than collapse.

    From here

    > Pop history has a nasty tendency to compress all of that into one idea of ‘Rome,’ which rises once and falls once, as opposed to the reality of a Rome which rose, fell into civil war, then rose some more, then had a crisis, then stabilized, then fragmented, then fell in some places while remaining stable in others. And so, for example, Sallust’s complaints about Roman decadence – which date to the first century B.C. nearly five centuries before its ‘fall‘ – are often quoted as somehow explaining Rome’s eventual demise, but Rome wasn’t even done expanding at that point.

  4. 4gravitons Says:

    The UC system is still using AP tests, IB tests, grades, etc., yes?

    I agree the SAT and ACT are more objective than essays, but I don’t think they favor the sorts of non-conformist smart people you’re imagining. I remember studying for the SAT being an exercise in conformism. For the Reading/Writing and Language sections, the task was usually not to pick the correct answer (which was often not even present) but to pick what you thought the test-writers believed was the correct answer. For the Math, it was not about knowledge, depth of thought, or carefulness, but the ability to guess answers to rote problems quickly. And the Essay was the most formulaic five-paragraph nonsense I’ve ever had to write.

    I’d hope that, in dropping the SAT and ACT, UC schools will instead prioritize the actual things the students have learned. Maybe that’s overly naive though.

  5. David Karger Says:

    Scott, since you are looking for optimism, here are a pair of predictions for you.

    1) 3 more months of watching other countries celebrate their covid-free normalcy and comparing it to the ongoing disaster of the US response (especially when those other countries ban US tourists to protect themselves) is finally going to strip away (just barely) enough support from Trump to let Democrats take the presidency and the senate. Having done so, Democrats will finally give up on their decades of trying to appease the Republicans and ram through some decent tax and health care policies in an effort to make us look more like those other countries that are doing so much better.

    2) Colleges are on the whole committed to a left-leaning, education-as-a-social-good model. Although they have now lost the SAT as an “excuse” they can use to admit the under-privileged, they will have no trouble finding different excuses. They’ll simply adjust their “shifting and nebulous criteria” to admit the same people they’re admitting now. The right will be shocked and furious to discover that the SAT was never *forcing* colleges to admit minorities, and the left will be shocked and furious to discover that the colleges *still* won’t admit minority students who are under-qualified.

    If you ask me how confident I am about these two predictions, I’m afraid I’ll only take about 50-50 odds on number 1 (so I’m certainly worried, but not depressed most of the time) and about 90-10 on number 2 (in the sense that I expect about 90% of colleges to go the way I predict).

  6. Boaz Barak Says:

    Hi Scott
    It’s not something that I studied carefully so don’t have a strong opinion about, but in the hope it makes you fee better, I believe the young you would still have gotten into top universities even without the SAT. If I recall, you graduated high school at 14 and already had an academic paper. The admission officers didn’t need the SAT to know that you’re a star.
    Generally standardized tests such as SAT are not very useful to distinguish between the top 1% and top 0.1% students.
    (This is not to say that this decision is good or bad – but just that it should be judged by its effect on the majority of applicants and not on the few academic superstars)


  7. Bunsen Burner Says:

    “So, what can anyone say that will make me less depressed?”

    Nope, sorry I’m afraid you are completely correct. The US political class actually gave up on creating a robust, resilient, and fair society a long time ago. Trump is very much a symptom rather than some new causal element. It would be good if US citizens started to think seriously about the type of society they want to live in without juvenile attempts to label anything they don’t understand as communism, but I am not optimistic about this. Given the disaster that COVID has wrought, I have zero expectation of the US surviving the crises brought about by Climate Change. May I suggest a browse through Joseph Tainter’s “The Collapse of Complex Societies” to understand something of the mechanisms at work?

  8. Russell Impagliazzo Says:

    I think you should include the SAT in the list of nebulous criteria that favor the rich and well-advised. The scores for students applying to UC are so clustered at the top that the information in these scores is almost negligible. As others have pointed out, what the SAT tests is rote learning and the ability to play the game, not any actual intellectual skill or accomplishment.

    A few months ago, I just read a whole bunch of UC students applications to pick some for a scholarship. (No, I shouldn’t have been given this job, and there was lots of randomness in the process, but that’s for another post.). The “leadership” part of the essays was dominated by robotics clubs, debate club, tutoring, science fair projects, volunteer work and being part of programming competition teams. There are plenty of ways for nerdy students to excel in these categories.

  9. Jon Awbrey Says:


    The ETS of today is a very different service with a very different agenda from the one, I’m guessing, both of us once knew. You could get a useful perspective on the current state of democracy and public education and many related things by following Diane Ravitch’s blog.



  10. ira Says:


    Population ~ 95 million
    GDP/capita ~$8000 (PPP)
    Land border with China
    327 Cases covid19
    ZERO covid19 deaths

  11. ryan williams Says:

    I did pretty confoundingly badly on the SAT. Took it twice. It cost my family a good chunk of change. (I did better on the ACT, but many schools didn’t accept the ACT.) I had some cheap SAT study books but none of them seemed helpful.

    It was probably my science fair antics that got me into Cornell (but not many other places). I think the SAT is good for identifying some people, but definitely not everyone.

  12. Robin Hanson Says:

    These latest bad things that have happened, they don’t seem to have much to do with Trump. Trump has been consistently trying to cut regulations, and he’d probably oppose cutting SAT from admissions.

  13. Alan Aspuru-Guzik Says:

    Hi Scott!

    I agree with your statements about the downfall of the US. The signs for all of this were brewing for very long we’re sparked for me with the tea party and the terrible behavior of the Republican-controlled house and Senate under Obama but hark of course way before then.

    Perhaps in the SAT front, I urge caution. It seems UC may create their own test instead of the racket of the SAT. My undergrad university UNAM had an extremely hard admissions exam that they designed themselves. Canada’s admissions as far as I can tell is based on high-school grades (marks as they call them here), which I find fair too.

    I am also in favor of objective admissions and maybe UC will move towards that, to early to tell.Other than that I commiserate and have nothing to share to make you more hopeful as I am not hopeful for America.

  14. Paul Topping Says:

    I am somewhat receptive to the idea that the SAT costs too much, it doesn’t accurately predict future educational performance, or that it favors students with certain backgrounds at the expense of others. Still, it concerns me that the powers that be are gradually eliminating all objective measurements of applicants’ intellectual abilities. It seems to me that this goes along with the general softening of our educational system. If I had to blame this on a single meme, it would be on the Left’s general desire to take the easy way out in their efforts to “fix” the system for the truly disadvantaged in our society. The sentiment is a good one but they are bringing the educational level down for everyone rather than fixing any problems.

  15. Shmi Says:

    > I feel certain that a 15-year-old me wouldn’t stand a chance in the emerging regime—any more than nerdy Jewish kids did in the USSR of the 1980s, or the US of the 1920s.

    Well, “a 15-year-old me” was a “nerdy Jewish kid” in the USSR in the late 1970s, and the situation was deplorable but not nearly as dire.

    While I and some of my classmates who were not on the unofficial “Schindler’s list” of university admissions had to settle for higher education in “non-Jew-restricted” universities and degrees, the culture of knowledge and respect for intelligence was as strong as ever. Science and math Olympiads helped pick the top talent, which was shepherded to the best schools regardless of ethnicity. Every university had oral and/or written entrance exams to select their cohort (sadly, the exams were also conveniently used to reject the undesirable Jews who dared to apply without being cleared ahead of time). But at least there were always reasonable choices. I was one of the “victims”, and got my degree in Computer Engineering instead of Physics. Some of my classmates who were similarly rejected still did very well for themselves, one was a prof at Tufts for a long time, another is an inventor and a business person in Israel, etc.

    Incidentally, the “Jewish restrictions” all but disappeared by the early 1980s, not that it helped Soviet Union all that much.

    Sadly, there is no similar culture of knowledge in the US, as far as I can tell. Tribalism is much more important, with various clusters on the spectrum between alt-right and woke-left. And so a 1%-fatality pandemic becomes a reason to fight, not to come together…

  16. static Says:

    Maybe the opportunity there is to replace the SAT / ACT with a more difficult test that allows for more separation at the higher end and coverage of specific subjects, like the AP tests, but with more score levels. These definitely favor non-conformists. Most of my teachers hated me, but couldn’t ignore that I had a score above the average high school senior when I was 11.

    More likely they will replace it with an easier test where a significant number of students get a very high score.

    However, the possibilities of remote education are great. If you could work to ensure that real college course with real college credit are available remotely to students that don’t yet have a high school diploma, kids could continue to progress and get access to the materials with more structure and incentive than the average online course.

  17. Hyman Rosen Says:

    Liberals push on ropes. That is, they believe that disparate outcomes are caused by racism *at the point of disparity* and that the problem can be addressed and fixed there. This is obviously wrong, and that’s why so many well-meaning social programs fail to deliver any improvements.

  18. Scott Says:

    Anon #1: Isn’t everything you said, about the possibilities of using money to game the SAT, like a hundred times truer for all the other criteria used in undergrad admissions?

    Money and connections are regularly used to game grades (send your kid to a fancy private school that gives out gentleman’s As), to game extracurricular activities and “life-defining experiences” (obvious), to game personal essays (hire a consultant to ghostwrite them).

    And yet, when Lori Loughlin and all those other rich parents were caught in the cheating scandal a couple years ago, it turned out that they’d outright fabricated their kids’ SAT scores. Somehow, even with all their wealth and power, they still found the SATs themselves too hard to game!

  19. Scott Says:

    4gravitons #4: So let me state my claim carefully. It would indeed be absurd to say that the SAT selects for non-conformity. What it does do, however, is give sufficiently smart non-conforming kids a fighting chance, by making it harder to justify rejecting them. In other words: just by being a test with clear right answers, one that won’t outright punish you for non-conformity, it somewhat counteracts the massive conformity-favoring bias of the entire rest of the system.

  20. fred Says:

    I’m really sorry that your dreams of American exceptionalism have been shattered.

    The true rank of the USA is actually somewhere between Switzerland and Ireland… according to the covid death rate *per capita*, the only true measure of the quality of a country!

  21. Gabriel Says:

    I see that in the US there are silly arguments about whether wearing a mask is “virtue signaling”. Very sad. Wearing masks can reduce both the risk of infecting others and of getting infected yourself. And even if you do get infected, you might get a milder disease. Granted, it’s not 100%, but nothing in life is 100%. The problem with wearing masks is that if no one around you does it, you feel kind of stupid wearing a mask yourself. That’s why everyone who wears a mask is helping others do so as well, and everyone who doesn’t is making it harder for others to do so. I feel fortunate that I live in a country (Israel) in which wearing masks is the law, and which successfully contained the virus by taking early decisive action.

  22. Scott Says:

    Boaz Barak #6:

      It’s not something that I studied carefully so don’t have a strong opinion about, but in the hope it makes you fee better, I believe the young you would still have gotten into top universities even without the SAT. If I recall, you graduated high school at 14 and already had an academic paper. The admission officers didn’t need the SAT to know that you’re a star.

    Awww, thanks! 🙂

    Just to set the record straight, I was 15, although I did have a published paper. I was admitted to Cornell and Carnegie Mellon and rejected literally everywhere else. Of course I can’t rerun the experiment, but it seems doubtful that the “skin-of-my-teeth” Cornell and CMU admittances (for which I remain grateful!) would’ve happened without the SATs.

  23. fred Says:

    The elephant in the room is that about 50% of STEM grad students in the US are (were?) coming from China (kids of rich CCP members).

  24. Alex K Says:

    What are you doing based on your predictions? Getting dual citizenship? Not bothering to save for retirement? Learning combat skills and wilderness survival?

    I’m not asking this sarcastically. I remember being a college student outraged by the failure to respond to the flooding of New Orleans by hurricane Katrina. But back then I blamed G. W. Bush – I thought the system was fundamentally sound and the problem was due to a lack of good leadership that could be reversed. Now I see the same failure taking place on a national and an international level (not only the USA is botching its coronavirus response). And I don’t think Trump’s to blame: sure, he’s part of the problem, but he’s not an aberration (or no longer an aberration) – this is how the system works, or fails to work, now.

    With that said, I’m a also a nerd who spends all his time in front of a computer; I’m dependent on the system that, as we see now, is simply not able to respond appropriately to crises. Consider that the coronavirus is relatively mild. What if it was like MERS, with a 35% fatality rate? I seriously think social order would collapse. I don’t want to move somewhere remote and stockpile food and ammunition (and there’s no way I would be able to convince my friends and family to go along) but maybe that’s the rational course of action?

  25. Scott Says:

    Robin Hanson #12: I agree! Trump is not the source of all evil in the universe, just like the black hole at the center of the galaxy is not the source of all gravity in the universe: it’s merely very, very, very massive. 😀

    If we compare the US to “peer countries” like Canada, it does seem likely to me that at least half of the American covid deaths can be directly attributed to the staggering incompetence of Trump and people who he appointed. But yes, “the collapse of the Leviathan” predates Trump and is larger than him; Trump is merely a (gigantic, horrible) symptom. And what depresses me the most are the people who despise Trump as much as I do, yet who seem determined to mirror his magical thinking on the left.

  26. orthonormal Says:

    Re: the latter, it may be a consolation to you (societally rather than personally) that I expect the majority of US universities to contract or fail in the next decade or two.

    Expensive online college this fall will be a farce, because it’s lacking the networking that’s the only value-add over a standardized online class with decent incentives. Thus I expect there will be startups asking students to take a gap year in 2020-21 and take their program in the meantime for an extra dot on their resume. (Most obviously for CS, where coding bootcamps are already better than an undergrad degree for knowing how to code at a job.) The program will be much cheaper, or even paid for by a cut of first-serious-job money.

    A decoupling of networking from learning and credentialing, and the normalizing of alternate credentials, spell financial doom for all but the most prestigious universities pretty soon. Tenure may cease to protect faculty, first for small liberal arts colleges, and then more broadly. (During this time, of course, the academic job market will collapse completely as it’s hard for a recent PhD to compete with an established researcher. The smart grads who used to become professors will go into Google’s AI lab.)

    If the startups go by the model of taking a cut of post-graduation income, then those startups will do better if they use standardized measures of applicant quality rather than nebulous ones. (Maybe with the exception of business, where having connections is the best way to earn big immediately.)

    So standardized testing may outlive the “everyone in the middle class goes to college” phenomenon.

  27. Chris Says:

    Here’s a way of looking at the SAT thing: though this (and a whole bunch of other things), suggests the decline of our higher learning institutes; the rise of very cheap ONLINE learning is more than making up for that.

    Just the other day you posted quantum computing lecture notes, and I read over them and they are excellent! Anyone from anywhere can decide to read them and learn about quantum computing.

    Why would they need to be admitted into a university in order to read these notes? Or to do the problem sets?

    thus if you think about it, with respect to accessibility of education, we are INCREASING it despite rising tuition costs, and silly “leadership” criteria.

    BTW: WHY THE HELL DO WE THINK HIGH SCHOOLERS SHOULD BE LEADERS? any high schooler with a modicum of self awareness should recognize: you should learn a lot more and gain much more experience before you become a leader. And yet, our University admission system forces high schoolers to pretend to be leaders.

  28. C Jones Says:

    Since in
    |Doofosity, Self-Referential, Fate, Feeling>
    the below is what I would hope a friend who measured in
    would find to be an “encouragement operator” to end in
    More and more it seems just another cognitive illusion that when I think I may have finally worked my way into the clearsightedness of the third quartile of the Dunning-Kruger chart I only discover that I was actually not on the Dunning-Kruger curve I thought I was and I’m back in the first quartile of doofuses who misjudge their own competence. Lesson: we are all doofuses on some curve that is (it seems) always hidden to us …
    On the testing question: Our best hope (and you remind us of this all the time so I remind you of it) is the scientific hope that it is possible to find some answer to well formed questions. Yes, questions and measurement bases applied to evolving systems (that may exhibit expanding state spaces) are always contextual, biased, or incomplete. But that just means there will always be work to do. The good news about the testing issue is it would seem to be the kind of problem that could be explored with well-formed questions, careful measurements, ongoing questioning of those questions and measurements and, more broadly, experimentation with different protocols in different states and institutions. Finding a measurement operator (in this case admissions criteria) is, you might agree, in general, a very hard problem. So, let me throw it back at you, as a decision problem, what can computational complexity offer to help the California BoR find a good/better/best measurement operator to apply to teenagers?
    The “Leviathan” collapsing: On the one hand I’m very sympathetic to my own impulse to nod along to declinism or catastrophizing. (Just ask my doctors). I mean … realistically … we are all TRULY doomed both in the not too distant future of our own lives and in the universal sense … And if I start to list how bad it all is then I’ll never get done. But … we can see there’s a long history of a tendency in us both individually and as cultures/countries to engage in a declinism bias. The “cure” for that is to see that well first of all it’s a particular bias or perspective, and thus, possible to reframe. That doesn’t mean that it isn’t STILL a “bad” or “harmful” thing (that is a perfectly “true” or “valid” measurement) but that it may be seen as well as other things like “motivation for change”, “learning”, etc. as others noted above. Again, to turn it back to computational things … it seems some of the thinking/motivation behind the US Constitution was how to avoid the historical fate of political systems declining or self-destructing and designing a self-correcting “control system” for government. The good news is that the evidence (so far (which is what we mean by evidence)) does suggest that
    1. Human nature being what it is there will always be evil, badness, doofosity, mistakes, etc., but …
    2. We STILL have a system in place that has the potential of being used to self-correct those evils, mistakes, and removing doofuses.
    because there is ample evidence from our own history of evils, mistakes, and doofuses being corrected by this control system. But, just like the example above, it requires actual work being done. That work will be done so long as there are those who care about doing it. And I am grateful and encouraged that there are (many) such as yourself who are caring and doing good work.
    Hopefully you find the good work encouraging and end up if not in |0011> then at least in some a|1100> + b|0011>

  29. fred Says:

    “If we compare the US to “peer countries” like Canada”

    Fact – Canada’s population is 11% of the US population.
    The USA has a much more varied and wider distribution of states, from a geographic, political, economical, and demographic point of view.

    Covid deaths per million capita:

    Canada (38M): 174
    USA (330M): 302

    When you look at neighbor border states/provinces (West to East)

    British Columbia (5M) 30
    Alberta (4.3M) 30
    Saskatchewan (1.1M) <10
    Manitoba (1.3M) <10
    Ontario (14M) 160
    Quebec (8.4M) 500

    Washington (7.6M) 158
    Montana (1M) 15
    North Dakota (0.7M) 71
    Minnesota (5.6M) 159
    Wisconsin (5.8M) 88
    Michigan (10M) 525
    New York (19.5M) 1507 (of course here the big outlier is due to NYC… I guess De Blasio must be an incompetent murderous idiot?)
    Vermont (0.6M) 87
    Maine (1.3M) 59

    It's clearly all over the place in both cases with a pattern that things are worse on the East side… who in his right mind wouldn't be surprised by this?
    Yes, things aren't as simple that one can blame the "mass deaths" (oy vey…) on Trump alone.

  30. Nick Says:

    Yesterday an unarmed Black man was murdered in the street by a White cop in Minneapolis. So it looks like things are finally getting back to normal!

  31. fred Says:

    As Cuomo just put it

    “government this, government that… forget government! This whole trajectory is decided by people, it’s personal behavior!
    There is a certain amount of informing the public and accepting a new type of standard… you put a mask on!”

    That’s really the main and unquestionable jab about Trump, the only one that matters, he won’t wear a mask in public…

  32. Ed Says:

    I won’t address all the various fascinating culture-war-adjacent issues (like many, I have strong feelings.)

    But I do want to “kick-the-tires” on what seems like a major point in the post, that “most chances of social advancement in the U.S.” hinge on attending a highly-selective college.

    What is this “social advancement” being mentioned? Why should we be promoting it? And what’s the evidence that a rejection from (e.g.) Berkeley impedes it?

    I probably seem really pedantic here, but I think many people’s values/intuitions are less aligned on these questions than is often appreciated, so clarification is useful.

  33. matt Says:

    Want to be less depressed? Just remind yourself that not long ago, quantum information science disproved the most important open conjecture in operator algebras (Connes’ embedding). I mean, wow! That is insane! I wonder what von Neumann would have thought.

  34. Art Says:

    Lots of comments responding directly to the points made above. Seems unlikely to improve your mood – so I’ll try something different.

    The American public agrees that schools should stay closed, stay at home orders are reasonable policy, and that we should continue social distancing for as long as required to deal with the crisis. All of these with a greater majority than those who like apple pie. Further, pretty much everyone is at least attempting the social distancing thing.

    I think that’s reassuring. The structures bad, but at least it isn’t emblematic of citizens themselves. As a bonus, we can expect infectious disease in general to take a hit after this is all through. Might even get a congress that can meet virtually, so lobbying will require actually being in all 50 states.

    In another direction, a CS fact that made me happy recently:

    There’s no (Turing complete) model of computation that has exactly half of all programs of length n halt on the 0 input.

  35. John Cherniavsky Says:

    For your depression why don’t you blog on Mahadev’s Quantum verification result.

    You know Scott there have been studies on the correlation between test scores (SAT or ACT) and success in college – what the studies showed was a small correlation for the first semester or two as measured by college grades. They don’t predict future success or graduation or any other realistic outcome – they predict that you can do well on these types of tests. Guess who benefits most from using the SAT or ACT – the rich kids who can afford tutoring which does work assuming a motivated student – a standard deviation improvement is certainly possible (100 points on the SAT). I guess the university needs to fall back on grades and AP tests and other means of judging which students they will admit.

  36. Paul Beame Says:

    I have mixed feelings about the SAT/ACT and I can see why one might not want to make them required but eliminating them altogether in admissions seems a bad idea. The SAT/ACT controversy and the need for them points out bigger issues.

    The entire college admissions process in the US is broken for a lot of reasons. A big issue in making judgements based on high school education in the US is the complete balkanization of the curriculum across (tens of) thousands of school districts and between schools.

    I grew up in Ontario, Canada where there was a fixed provincial curriculum for all schools – with advisory committees for each course made of high school teachers and discipline-based university faculty – so that a course of a given year and level at one school had the same content as another, modulo a few options to be covered – though exact grading rigor might differ between schools. (There was also no political interference like the Texas board to approve texts and ruin biology education for an entire country.) These curricula were sufficiently uniform that one could apply to any one of a ten universities in the province with a single 2-page application form that worked for all of them, though University of Toronto had a subsidiary college system that required an additional 2-page form for admission.

    It took an hour to fill out everything! What a far cry from the months of misery my kids had in filling out applications! (I did take the SAT and some AP tests to apply in the US because my parents wanted me to apply there. I was a really good test-taker and high school math contests had honed my multiple choice skills. I got in, but decided that paying many times the amount to go to an ivy league school just wasn’t worth it. I haven’t regretted that decision.)

    I don’t find the SAT/ACT or ETS inherently problematic. I don’t even think that the content of either of these tests is necessarily particularly problematic. There is a place for testing basics in reading comprehension and basic reasoning or math skills. These do seem appropriate for an SAT/ACT if you can’t judge those based on high school grades.
    However the whole system of test preparation courses that makes companies like Kaplan the big bucks is a corrupt one skewing the system away from being fair, at least wrt socio-economic factors.

    In the 2000’s I spent some time on the CS GRE subject test committee at ETS. (I had taken the Math subject test back when I applied to grad school at a time when it was really trivial but never took the CS test.) I was impressed with the care with which the ETS made sure that one could check the validity of questions and equate the scores from different sittings. It matters exactly what the total difficulty of the questions is, when in the sequence a question is asked, how well a question correlates with doing well on the exam overall, etc…many more factors than I knew existed.

    However, I also quickly got convinced that the CS subject tests themselves were a stupid way to test computer science undergraduate achievement, and told as many of our undergrads as possible not to take them. While things like reading comprehension and basic reasoning are well tested on standardized tests, and there were some parts that involved reasoning on the CS subject test testing some particular jargon seemed particularly stupid. (The last universities requiring the CS subject GRE before it was eliminated were weak undergrad institutions using it as an exit requirement.)

    I don’t get to see undergrad admissions, but for our grad applicants the GRE does serve some purpose, even though letters from people we trust and undergraduate record are much more important. GRE Quantitative scores are almost always high, so that is almost never an issue, GRE Verbal scores are useless for those applying from non-English speaking countries, and are of marginal utility, I generally only find them a significant negative for those applying from the US with really bad scores; on the other hand, I suspect that the GRE Analytical writing scores are clearly saying something for English speakers: If someone from the US is getting a 2.5 or below, then that seems a bad sign no matter what the rest of their GRE scores are, and similarly if they are getting 4.5 or above, that seems a good sign.

    Of course in the current environment, this isn’t required…

  37. Scott Says:

    ryan williams #11: Your story is an inspiration to me, and I’m proud to have played a tiny part in it, starting when we were both undergrads. I’m surprised about the SAT but glad that you were able to get to Cornell on the strength of a science fair. (My project didn’t qualify for anything at the Westinghouse competition, but was then accepted for publication at ACM SIGIR.) As I said the last time this was debated on Shtetl-Optimized, my ideal admissions system would be a Boolean OR, over super-high test scores or a super project or overcoming of super adversity in life, with the bar for each of those things simply raised until one has the right number of admits.

  38. Rahul Says:

    About being less depressed:

    Have you watched war and peace especially the BBC 20 episode series? That one somehow gets me out of my depression. There’s just something about war and peace that lets me get perspective.

    This too shall pass.

  39. Raoul Ohio Says:


    This should make anyone less depressed: online access to math and science seminars ALL OVER:

    Representation and propagation of uncertainty for me! I’m sure about it!

  40. Scott Says:

    John Cherniavsky #35: I already blogged about Urmila’s breakthrough (google for it).

    What you write is at variance with all the research I read, which said that the SAT is about as good a predictor of future outcomes as you can find (certainly SAT+grades improves over the predictive power of grades alone). At any rate, that was the conclusion of the recent report by UC’s faculty senate, which Janet Napolitano and the regents set aside because it wasn’t what they wanted to hear.

  41. John Cherniavsky Says:

    Here’s a link to Phil Sadler and colleagues work on the AP exam – I thought he also worked on the SAT. – but I was wrong. He’s been at Harvard for a while and that is not a typical university but he worked with colleagues at less elite schools. He was the guy who interviewed Harvard graduates at graduation time and asked them questions like “Why does it get hot in the summer?” Typical answers were – because the sun is closer (which the students should have immediately realized was wrong!) Astrophysicist and Science Education researcher.


  42. Stephen Jordan Says:

    I’m sorry to hear you are depressed, Scott. I don’t disagree with your assessment. My advice would be to attempt to avoid allowing your emotions to become tied to things over which you do not have personal control.

    It is very natural to think that if vast and terrible things are happening in the world then we should of course devote a lot of attention to learning about their details and figuring out how they should be dealt with. However, I believe I should only devote time to this if either:

    a) I find it fun


    b) There is a non-negligible likelihood of me actually having some influence

    I won’t presume to advise on how these criteria apply for you in specific cases, because it is not obvious. (For example, maybe a compelling blog post by a prominent academic such as yourself could help preserve standardized testing.)

    But for me framing things in an explicit consequentialist-utilitarian way is surprisingly helpful. Some might consider it too obvious and trivial to repeat, but I nevertheless find value in repeating to myself that the purpose of emotion is to drive action. And mourning or ruminating over the terrible decisions of people over whom I have no influence is of no more value than cursing the rocks after an earthquake.

  43. Andrei Says:

    Why don’t universities give admission exams? Each one to have its own exam, not standard stuff like SAT. You want to study CS at Harvard, you show up at Harvard on July 22 when the Harvard CS exam is scheduled and take it. First 100/1000/whatever (no idea how many CS students are enrolled each year at Harvard) are in, the others are out. (That’s roughly how stuff works in Eastern Europe.)

  44. Scott Says:

    Andrei #43: Sure, that could also be a good system (although in the US, students typically apply to 10-20 universities, which would translate into a lot of exams).

  45. Scott Says:

    matt #33:

      Want to be less depressed? Just remind yourself that not long ago, quantum information science disproved the most important open conjecture in operator algebras (Connes’ embedding). I mean, wow! That is insane! I wonder what von Neumann would have thought.

    Thanks; I’ll remind myself of that more. 🙂

  46. Sniffnoy Says:

    People are making a lot of the same argument over and over here — that the SAT favors the wealthy because of test prep courses.

    This argument is, AFAICT, mostly wrong; and to the extent that it is correct, it’s not correct in a relevant fashion, in that the bias is less than for other methods. (Scott has already made that latter point, but I think it’s worth repeating.)

    John Cherniavsky #35 claims such courses can improve such scores by 100 points. John, do you have anything to back up that claim? All the numbers I’ve seen suggest that average improvements are maybe only about 20 points. Here’s an article about this, with some links to sources.

    So, this supposed test prep bias in fact appears to be tiny. This shouldn’t be surprising; the SAT is roughly an IQ test, and basically nobody has found any way to consistently improve performance on those.

    Of course, it’s not zero. But this brings us to the other point, the one Scott already made — the key question isn’t, is it biased, it’s, how does its bias compare to that of other methods? And in particular, how does it compare to that of the specific other methods that will be used in place of it?

    And the answer there seems to be that it compares pretty damn well. As already mentioned, the test-prep bias seems to be quite small. (And note that this is implicitly granting the premise that test-prep is pure test preparation that only increases test scores and doesn’t at all actually make one better at what’s supposed to be tested!) Whereas the infamous extracurriculars and essays seem to favor the wealthy or the well-connected to a much greater degree. Or they favor those who one might euphemistically call the “socially adept” — the manipulators and bullshitters and bullies and etc.

    One can pay for a ghostwritten essay. Looking for leadership and extracurriculars obviously favor politickers and bullshitters, and having the resources (not only time but also money, depending) for such things can depend on familial wealth. How much? Well, that’s hard to measure, but it sure seems to be a lot more than the measly 20 points one can maybe add to one’s SAT score with test prep.

    I mean, Scott already pointed it out — in the recent cheating scandal, SAT scores were outright fabricated! If the SAT were so biased, there would have been no need for this. I mean, did people, like, blatantly make up extracurriculars in the same way? Having an essay ghostwritten is admittedly a similar sort of fakery, but since authorship isn’t checked there, that’s far easier to do, making it once again more biased than the SAT.

    Or, look — when Harvard and other such school switched from a purely exam-based system to a system based on “character” — which was, recall, deliberately intended to keep out the Jews, who were doing so well on the exams — do you think they failed at their goal? Do you think their deliberate attempt to put their thumb on the scales actually made things accidentally fairer? No, of course it didn’t! If it had they would have immediately gone back on it.

    In crude terms, tests relatively favor the Nerds, while extracurriculars and essays relatively favor the Suits! The choice should be clear.

    Hell, the nature of the test doesn’t even matter as much as you might think. Harvard’s introduction of “character” evaluations was in 1926; their entrance exams at the time weren’t the SAT (they adopted the SAT in 1935). Or for an extreme case, as Razib Khan points out, the old Chinese imperial examination system — which, AIUI, used exams that were basically nothing but memorization, not at all the sort of test we would like, nothing like the fairly preparation-immune SAT — still functioned to promote those without connections into the bureaucracy, who’d never have had a chance otherwise!

    Another argument against the SAT or GRE that is occasionally brought up is that these tests can’t be any good because they don’t predict future performance. I don’t think this is true; a number of studies drawn on students from a fixed school have found this, but the selection bias makes these results meaningless. (Remember: Within the NBA, height has little correlation with performance! Doesn’t mean height doesn’t matter to basketball — selection processes will do that.) I don’t think this is true of the overall population. Here’s what the College Board themselves have to say about it — the correlation appears to be pretty good as such things go (nowhere near perfect, but again, that’s not the question).

    No, tests, the SAT included, aren’t perfect. They won’t successfully find every capable person who otherwise would be overlooked due to lack of connections or social graces. But having tests at all does much more for that purpose than not having tests; and indeed the SAT specifically is actually a pretty good test, where preparation courses don’t seem to help that much.

    People have been making these same arguments a lot even though Scott has already pointed out some of the reasons why they’re not true. It’d be nice if people would actually respond to and argue with this, instead of just repeating the same mistaken argument over and over.

  47. Sniffnoy Says:

    Andrei #43: That’s how things used to be over a century ago. Then they decided to get together and standardize things, which is where the SAT and the College Board came from.

  48. Paul Beame Says:

    Scott (#44): But US colleges already *do* in effect have separate entrance “exams”, in the form of application questions designed to elicit exactly what it is about their institution that makes them special to the applicant, or off-the-wall questions like the notorious University of Chicago one each year (which I actually think was the most interesting).

    This is just one of the reasons that the “Common App” is such a misnomer. When my kids were applying, no two schools had close to the same requirements with the Common App. The UC schools didn’t use it and used their own application. A bunch of schools now use the “Coalition App”, which I hear is a major improvement over the Common App and way easier for low-resourced people to fill out, but of course the UC system isn’t part of it. Which would be more important for socio-economically disadvantaged students: dropping the SAT/ACT, or joining the Coalition App and reducing the barrier for students to apply to a variety of colleges?

  49. Sniffnoy Says:

    Btw, I’d like to make a follow-up to my comment #46.

    A lot of people have framed this in terms of rich-vs-poor. But well-connected vs not has come up too, and I think that’s closer to what’s really relevant. Because the real question isn’t rich-vs-poor; it’s about people who are good at actually doing things, vs people who are good at fooling people. The Nerds and the Suits, as I said. That’s the real scary bias introduced by these tests of “character”.

  50. E. Harding Says:

    “So, what can anyone say that will make me less depressed?”

    Montana seems to have succeeded, I guess. Florida’s epidemic has been far, far less severe than that in Massachusetts. Peak case counts always postdate the actual peak of new infections by weeks. China, the country with the first large COVID outbreak, has so far held its Two Sessions without issue and has tested the majority of Wuhan with only a small cluster being found.

    Hanson once again demonstrates himself to be a troll.

  51. Anonymous Says:

    I think if you are someone who trusts that admissions officers know what they’re doing, then this is not as bad as it seems. To me, it seems easier to admit a student body that is non-conformist by looking at extracurricular activities, essays, or the classes they choose to take, than it does by looking at SAT and ACT scores. But this assumes that admissions officers can distinguish non-comformists from those who are just merely well-connected on the basis of these factors. I’m confident in their abilities, though I can see why not everyone is.

    One thing that doesn’t seem to be brought up in these discussions is that by requiring standardized tests, it sends a message that you should spend a significant amount of time preparing for them. On the other hand, I think many colleges don’t want to send this message, for good reason. I think many applicants to top schools are much better off pursuing activities they’re interested in, like say reading about quantum computing, than studying for the SAT or ACT, and I hope the admissions officers at these schools agree. And I think the easiest way to send this message is to just avoid considering them all together.

  52. Chris Blanchard Says:

    Hmm. I get angry, which shades into depression, but! The USA has been in a worse state than this several times this past hundred years. I am sixty-four, so I remember a bit of it, but starting before my time you had McCarthyism, which ruined many thousands of lives, and burnt books; and in lineal sucession you had Nixon (who was a House UnAmerican Activities Committee member and McCarthy’s ally), with a whole series of evils, and a landslide election victory to show just how many Americans were pro segregation, and the racist murders which went with it; and then Reagan (who’s political rise came out of that same set of ‘anti-communist’ lies, and who inherited Nixon’s party. And besides the mass murders both of them organised in other countries (later for Reagan of course), you had a considerable amount of blood on 1960’s and ’70s American streets, in big riots and with the National Guard and police forces murdering various sorts of American (as at Kent State for a small but visible example). Reagan, as Govenor of California responded to demonstrations by saying, “If it takes a bloodbath, let’s get it over with”. For a sense of the sixties, at least, it might be worth listening to Jefferson Airplane, particularly ‘Volunteers of America’. They, and a lot like them, couldn’t see any way out except to prepare for armed resistance. You might get there yet, and there are some Americans already think that way, but I don’t yet see it, so far.

  53. Ted Says:

    I have no idea if this fact will make you less depressed, but I’ll give it a shot:

    Millions of years of evolution took our ancestors to what I might call “chimpanzees++”: hominids with the very rudiments of human communication, tool use, and culture, but no capacity to exceed tight neurological limits. This is all well-explained by standard textbook evolutionary theory. But then – for reasons still unknown to modern science – in a relatively short time, evolution gave our brains an almost step-change upgrade that massively overshot the minimum system requirements for outcompeting our hominid compatriots. As far as I understand (while not being an anthropologist), the people of 70,000 years ago would have been statistically just as capable as we are today of understanding music, art, philosophy, existential despair, general relativity, and the Connes embedding conjecture, if they’d been brought up in a modern environment – despite the fact that their day jobs consisted primarily of foraging and not much else.

    And because of that upgrade, while there’s still an immense amount of both misery and idiocy in the world, a nontrivial fraction of the world population has at least the potential to experience the luxury of pondering oracle separations of BQP and PH instead of focusing entirely on subsistence, or of engaging in organized social activity that attempts to further improve the overall human condition – albeit through collective organization systems that are immensely imperfect, but still probably at least slightly better on expectation than purely random actions. And, of course, we can have the very meta experience of being confused about how evolution could get us to the point of being confused about how evolution works.

    (For the record, I’m not suggesting that this upgrade required any mechanism beyond naturalistic evolution. But it did require a mechanism beyond my very simplistic expectations of what evolution would produce – and we’re all very lucky that my naive mental model of evolution is incorrect.)

  54. Jalex Stark Says:

    My faith in humanity rests on
    1. the increasing concentration of power of competent non-state actors, e.g. Amazon, Microsoft, Google, OpenAI
    2. the fact that the interests of the above are mostly correlated with the interests of average people

  55. Steve E Says:

    Prior to starting college, I attended an ultra-orthodox Yeshiva, where there wasn’t much secular education and where most students didn’t go to college at all. In my high school, we weren’t taught history, physics, literature, and spent only a couple hours a day on secular studies altogether. I recognize that this makes me an aberration in the scheme of things, but given my background, my only path to university was doing well on the SATs and taking some classes at a SUNY school to establish some academic record. While I recognize it’s imperfect, I’m sad to see the SAT go as well, because I wouldn’t have been accepted to a university without it. FWIW, I attended Cornell too, and started college at 16.

  56. Rich Peterson Says:

    Every country humiliates itself for its ineptitude, somehow, usually frequently. We’re not a manufacturing nation anymore, so the lack of facemasks is lack of ordering from China, by the gov’t, because of over 20 years of republican leaders’ ineptitude, laziness, mismanagement, petty sabotage, and unwillingness to let dems do better than they would(I’m a Krugmanite). All we need to do is get out the vote, keep trying to do the right things, and support investigative journalism. I say this because in spite of the facemask fiasco, so many many things in usa have improved since Armstrong was on the moon-like minority rights, women’s rights, human rights, FDA(won’t see fda miss a anythin as bad as thalidomide), the fbi’s respect for the law now, and also,(this is progress)at least some eining of the cia and city police forces, remorse for our terrible crimes like to AfricanAmericans, Native Americans, standinf aside durung Holocaust.

  57. Gerard Says:

    Chris Blanchard #52

    I’m not convinced that any of what you describe really represents a worse state of society than what we now have. Yes, the 60’s were turbulent but at least there was an active opposition to the regime instead of today’s general apathy. I’m too young to remember much about Nixon but I think his domestic policies were in many ways more liberal than subsequent Republicans and for all his faults he at least had the decency to resign when confronted with impeachment. Can you imagine Trump doing that ?

    As for Reagan, I think in many ways his era was the start of the country’s slide into what it has become today, but it was by no means the nadir.

  58. randomdude Says:

    Your feelings about the SAT are too alarmist, these changes would have affected young Scott Aaronson, not kids with learning disabilities, or kids who barely know english or kids from third world countries who have a completely different HS curriculum, or kids with real resumes.

    Rich well connected kids are already able to go to whichever university they want to, ask Jared Kushner.

  59. Sabine Says:

    I went for a walk yesterday and a little boy, maybe 2 years old, carefully picked up a bug from the pavement and put it in a bush “so that no one steps on it”. Best I can do. Hope you feel better soon.

  60. K. Peterson Says:

    As an often-depressed person, I can tell you that looking for something outside of your control to relieve your depression is usually impossible. Look within to your family, your work, your social circle, hobbies, entertainment, self improvement. Everything else will just feed your feelings that everything is out of control.

    “When an insurance broker tells you that SF doesn’t deal with the Real World, when a chemistry freshman informs you that Science has disproved Myth, when a censor suppresses a book because it doesn’t fit the canons of Socialist Realism, and so forth, that’s not criticism; it’s bigotry. If it’s worth answering, the best answer is given by Tolkien, author, critic, and scholar. Yes, he said, fantasy is escapist, and that is its glory. If a soldier is imprisoned by the enemy, don’t we consider it his duty to escape? The moneylenders, the knownothings, the authoritarians have us all in prison; if we value the freedom of the mind and soul, if we’re partisans of liberty, then it’s our plain duty to escape, and to take as many people with us as we can.” – Ursula K. LeGuin

  61. Dandan Says:

    >So, what can anyone say that will make me less depressed?

    Depression is not an option now. There is a war ongoing. War on minds.

    You’re wrong when you say that smart are not in charge. Trump, Putin and their mob are very smart in what they are doing. They just have a different goals than you and other honest people have.

    They are building the tribal order. The order of power and obedience.
    The order where the truth is what a higher rank creature says and wants.
    The order where slavery is a norm of life.

    To achieve it they need to destroy the current free world order.
    They seed chaos, uncertainty, disbelief, injustice and hate.
    They split people, nations, countries, unions.

    This pandemic situation is actually good for them. Why fight it?

    So, what we should do?
    Spread the truth, be honest, keep our thoughts clear, fight fear and never lose our faith.

  62. Boaz Barak Says:

    Robin Hanson:
    I wouldn’t describe President Trump as having a principled “anti regulation “ stance (or as having any principled stance except “pro Trump”).

    To quote the latest Tweet:

    Republicans feel that Social Media Platforms totally silence conservatives voices. We will strongly regulate, or close them down, before we can ever allow this to happen.

  63. Andrei Says:

    Scott #44:

    Indeed that may be a problem in the US: how it gets solved here is that people generally choose beforehand a city to study in (generally the one closest to home) and apply to multiple study programs there (like 2-5 though not 20-30) including ones where they have a greater chance at succeeding (i.e. a soft option with less competition). A side effect of this is that people more frequently apply to different majors at the same university rather than the same major at different universities.

  64. 1Zer0 Says:

    Well with all the endless realms of MWI Multiverse, Eternal Inflation Multiverse, Stringtheory Landscape, Mathematical Universe Hypothesis or even Modal Realism it’s virtually guaranteed that there is a more “sane” copy of the world in existence, I find that highly encouraging. Accordingly, of course there is a (or infinitely many) universe where Trump is president and there is one where noone even knows that name.

  65. anon2 Says:

    As Fred indicated. The worst of the US Covid problem was from New York. DeBlasio and NY officials encouraged the chinese NY celebrations. February 5, NY Health Commissioner Barbot on Twitter, “Today our city is celebrating the Lunar New Year parade in Chinatown, a beautiful cultural tradition with a rich history in our city. I want to remind everyone to enjoy the parade and not change any plans due to misinformation spreading about coronavirus.” Feb 9 , Mayor Bill de Blasio and other officials jammed a stage on Hester Street on Mott Street — the heart of Chinatown.”
    “We know in China, so many of our loved ones are facing the challenges of the coronavirus, but we stand together,” de Blasio said. Senator Chuck Schumer was also in attendance, and he declared, “We love the fact that so many people come from all around the globe and make our city and our country a better place.” March 2, Deblasio twitter, “I’m encouraging New Yorkers to go on with your lives and get out on the town despite Coronavirus.” March 9, The mayor also elaborated on his reluctance to close city schools. March 12, Appearing on CNN, de Blasio offered some rare praise for the president: “I accept the notion of travel bans in this environment. (On Jan. 31, Trump travel ban announced effective Feb 2.) …. NY Gov Cuomo. In March, Cuomo’s administration issued an order that allowed nursing homes to readmit sick patients without testing them for Covid-19. Amid allegations of undercounted casualties, the governor also pushed back against pressure to have state regulators more stringently record and report death rates in nursing homes. NY Nursing homes had over 5,000 deaths.

    And then came Cuomo’s annual budget – which included a little-noticed passage shielding corporate officials who run New York hospitals, nursing homes and other healthcare facilities from liability for Covid-related deaths and injuries…. The NY Times recorded the exodus of people from NY seeded most of the US with coronavirus cases. Many went to Florida but Florida had few deaths than California. Florida has a Republican governor. … So how is the New York mass gatherings in February Trump’s fault. How should Trump have stopped Cuomo from screwing up Nursing homes in NY? Lots of seniors in Florida they did not screw up. Plenty of coronavirus cases coming from York to Florida. US case levels without New York look very similar to Canada.

  66. Anon Says:

    Agree with the comments that say that things like SAT, GRE, etc. are all formalized nonsense.

    I’ll give you a funny story just to keep the mood light. As a wannabe grad student (back in the day) from a foreign country, we had to take the TOEFL. (Never mind the fact that we’d all spoken English from birth!)

    The essay was “What is the favorite room in your house?” – this was hilarious at multiple levels. I remember doing a mental face palm when I saw the question. You’re asking late teens about rooms in their houses? What if they don’t have multiple rooms like rural Africa or something?

    So you just had to shovel them the shit that they wanted to hear.

    My friend whose family was rather wealthy hadn’t quite grasped this concept and proceeded to write about the “toilet” and how he enjoyed the process (!!!), liked to read there, etc.

    Needless to say, he flunked the test.

    I still crack up thinking about it all these years later.

  67. fred Says:

    “So, what can anyone say that will make me less depressed?”

    The words of Marcus Aurelius:

    Don’t let your imagination be crushed by life as a whole. Don’t try to picture everything bad that could possibly happen. Stick with the situation at hand, and ask, “Why is this so unbearable? Why can’t I endure it?” You’ll be embarrassed to answer.
    Then remind yourself that past and future have no power over you. Only the present – and even that can be minimized. Just mark off its limits. And if your mind tries to claim that it can’t hold out against that… well, then, heap shame upon it.

  68. Gerard Says:


    Could you recommend some resources (preferably free) to learn just enough about quantum computing to understand how to simulate quantum circuits with a #P oracle ?

    As a hint of where I currently am, as I understand it quantum circuit simulations are probably done mostly with matrix multiplication, the problem being that the matrices grow in size exponentially (as does the size of the state space of the circuit) as qbit and gate counts increase. What I don’t understand is what that has to do with counting problems.

  69. jonathan Says:

    Ironically, it was just this sense of civilizational decline that Trump exploited to become elected in the first place.

    I wish I could say something to cheer you up, but I’m afraid I don’t have much to offer. I actually think the situation is a good deal grimmer than you recognize. Our capacity to successfully handle challenges is collapsing even as the challenges we face mount.

    Perhaps the most hopeful thing I can say is this: all is not yet lost. There is yet hope while we have breath in our lungs. And while America may be in decline (and indeed the entire West), it is not yet gone, and we may yet fight against the dying of the light.

  70. Scott Says:

    Gerard #68: I’m convinced that, if I put up a post on this blog saying that quantum computing doesn’t matter anymore because nothing matters anymore, the entire world is collapsing all around us, and in fact, I’m not available to reply to comments because by the time you read the post, I’d already taken my own life … then one of the very first comments on that post would be “Scott, I’m confused about what it means to separate P from BQP ‘relative to an oracle.’ What’s an oracle? Could you explain it to me again?” 😀

    So, OK, BQP is in P#P because you can take each final amplitude in your computation, and write it as a sum of exponentially many contributions. In other words, you never write down the exponentially large vectors or matrices explicitly (since you don’t have room for that). Instead, you imagine “expanding out” all the matrix-vector multiplies into a gargantuan sum of products of entries. Then you simply evaluate that sum in #P. If you still don’t understand that, try the lecture notes for my grad class, or Vazirani’s lecture notes, or Quantum Computing Since Democritus.

  71. Gerard Says:

    Scott #70

    Thanks, that should be helpful.

    Maybe nothing matters, but I figure if we’ve been at it this long we might as well stick around to see how things turn out.

  72. Scott Says:

    Gerard #71: Right. Bertrand Russell famously said that, as a teenager, he was deterred from suicide only by the desire to learn more mathematics—words that resonated with me like few other words ever written by anyone. For indeed, when I reflect on it, no matter how bad things got (excluding situations involving, e.g., excruciating physical pain), for me being dead seems strictly dominated by staying alive in an infirm, zombielike state where I at least get to find out when one of my open problems is solved! That, and my kids, are two overwhelmingly obvious reasons to stay alive.

  73. Scott Says:

    David Karger #5:

      3 more months of watching other countries celebrate their covid-free normalcy and comparing it to the ongoing disaster of the US response (especially when those other countries ban US tourists to protect themselves) is finally going to strip away (just barely) enough support from Trump to let Democrats take the presidency and the senate.

    And here’s my terrifying counterprediction (which I’ll give at least 35% odds): you and I both think Trump loses the election. Trump, however, refuses to concede on some preposterous ground (“the Democrats bussed 10 million Mexicans to the polls!”), and the Supreme Court sides with him 5-4. He remains in the White House. Joe Biden eventually concedes, just like Al Gore did in 2000.

  74. Scott Says:

    1Zer0 #64:

      Well with all the endless realms of MWI Multiverse, Eternal Inflation Multiverse, Stringtheory Landscape, Mathematical Universe Hypothesis or even Modal Realism it’s virtually guaranteed that there is a more “sane” copy of the world in existence, I find that highly encouraging. Accordingly, of course there is a (or infinitely many) universe where Trump is president and there is one where noone even knows that name.

    Yeah, but should I be happy that all those better universes might exist, or sad that I don’t get to live in one of them? And conversely, regarding any universes even worse than ours: should I be sad that they exist, or happy that we avoided them?

  75. Scott Says:

    Art #34:

      There’s no (Turing complete) model of computation that has exactly half of all programs of length n halt on the 0 input.

    Well, that could happen for infinitely many values of n, but not for all of them (since otherwise the halting problem would be solvable in that model).

  76. questions Says:

    > Obama might be remembered by history as America’s last smart-person-in-charge, its last competent technocrat

    Google, whose administration did Janet Napolitano work for? Oh? Oh.

    Teach your kids how to write good Diversity Statements Scott. It’ll be the only thing that matters.

  77. kaathewise Says:

    Scott #75:

    > Well, that could happen for infinitely many values of n, but not for all of them (since otherwise the halting problem would be solvable in that model).

    Forgive me asking, but how knowing that it’s exactly half of the programs that halt would make the halting problem solvable?

  78. Scott Says:

    Ed #32:

      What is this “social advancement” being mentioned? Why should we be promoting it? And what’s the evidence that a rejection from (e.g.) Berkeley impedes it?

    Yeah, in my more optimistic moods, I sometimes muse: under a fully “holistic” admissions regime like the one we’re rapidly moving toward, I could never have become a CS professor, and would’ve had little choice but to become a superrich startup founder instead! 😀 The trouble is, even assuming I had any ability in that direction (a big assumption), the uncredentialed dropout who founds a Microsoft or Google or Facebook might also be a relic of a bygone age. After all, Silicon Valley now attracts even more fire than the universities do, from those clamoring for purges and for thoroughgoing reform.

    (It’s striking to me what fraction of political discourse feels laser-focussed on identifying, and then trying to shut down, each and every possible route by which someone might succeed in life despite a lack of social and political skills.)

  79. Scott Says:

    questions #76: That I consider Obama fundamentally smart and decent doesn’t imply that I agree with every view he’s expressed, let alone every view of everyone who worked in his administration.

  80. Scott Says:

    kaathewise #77:

      Forgive me asking, but how knowing that it’s exactly half of the programs that halt would make the halting problem solvable?

    Because you could run all the n-bit programs in parallel, keeping a count of how many of them had halted, until the counter reached 2n-1. Then you’d know that all the remaining programs ran forever.

  81. happy life Says:

    To have my kids, to see them growth, to watch them flourish and let me take part of their life… who cares a few monkeys making some noise?

  82. stephen matthew shannon Says:


    I desperately want you to feel better; partly because my outlook is “Bell-state” coupled to yours, but mostly because we need you as a healthy resource.

    If it makes you feel any better, then rest, at least partially (asymptotically) assured, that every living <> system that we know of; and quite a few somewhat less complex systems, self-organize (a self-reference to the self-referential position of this comment in the comment heirarchy), to a critical point (or attractor) that is a state of “order at the edge of chaos”. {{ yes, I purposely placed the period outside of the quote, because that is the well-formed logical position for it, despite editorial tradition }}. Economy (political economy) and ecology have the same logical foundation. These systems work the same way. Every living system drives itself to this point. And then it occillates in at least 1/f mode, and up and down the power-law spectrum, and frequently ventures into chaos (usually determined by hyper-connectivity of its network fabric), all the while self-generating its own adjacent space of possibilities. It’s natural. It’s evolution. It is the way it is. You are one of those people (yes, like Feynman) who are themselves right at the edge of the Void. This confusion you feel is normal: it is the emergent you, evolving, adapting, learning. This is what it feels like.

  83. Sniffnoy Says:

    Scott #80:

    I had assumed Art meant half in the sense of some sort of natural density, rather than half in the sense that for each length it’s half of them?

  84. Allemaraiccire Says:

    Scott, you have focused on a couple of minor peripheral issues, and if anything you should be much more anxious and depressed than you already are. China (really the Chinese Communist Party) has unilaterally launched a new Cold War on the world, which could quickly escalate to World War 3 at any time. The initial escape of the pathogen from the Wuhan lab may have been merely negligent and reckless, but then there were several weeks of coverups, preventing investigation (to this day), vacuuming up protective equipment from around the world, and deliberately spreading the virus and disinformation globally, all with the eager assistance of the WHO. Instead of making ammends for their attack on the world, they have become increasingly aggressive and belligerent, totally abandoning any pretense that they have ever had any intent to be a good global citizen, militarily threatening their neighbors (Hong Kong is gone, Taiwan is next), economically retaliating against anyone who criticizes them or even fails to lavishly praise them, and engaging in propaganda wars around the world. The Chinese Communist Party is global enemy number one, and this Cold War will continue until either (1) they have been destroyed, or (2) they take over the whole world bringing the entire human race to submission, or else (3) World War 3 happens. This is the only issue in the world that matters now, or to put it another way, all that matters about any issue is which direction it will push this cold/hot war that China has started. The CCP needs to be forcefully confronted and opposed by the USA together with most other countries. The USA is necessary for this, but not sufficient. The USA needs the leadership that will do this, and they have this leadership right now. Meanwhile, the Democrats are hellbent on destroying the economy, destroying borders, destroying democracy, rigging elections and subverting the real will of the people, abusing power with crooked institutions, and colluding with most media (including tech companies) who have abandoned news and opinion in favor of full-blown Goebbels style propaganda. It has already been clear for the past few years that if Trump doesn’t win re-election in November 2020, the Democrats will rig things so they never lose again, and it will be game over for the USA, not just the end of democracy, but the USA would collapse into a failed state. Now with China’s escalation, it is clear that they would love this Democrat-led self-destruction of the USA, as the CCP would then be unstoppable. Without another Trump term it would not only be game over for the USA, it would be game over for the entire human race.

  85. Anonymous Says:

    Scott #22

    It’s not clear to me why you think your SAT scores were the reason that you were accepted to Cornell and CMU. This implies that you feel like your SAT scores were the strongest part of your application. But for example, as Boaz mentioned, you did have a paper. To me, a paper is way more impressive than high test scores.

  86. Art Says:

    Scott #74
    > Well, that could happen for infinitely many values of n, but not for all of them (since otherwise the halting problem would be solvable in that model).

    Ah! You’re right. That’s cool. Do we know if any model actually satisfies this? (Your proof is the same as mine)

  87. Scott Says:

    Anonymous #85: One issue was, I only found out that my paper had been accepted at SIGIR the week after I got a huge pile of rejection letters from universities. (Literally! I remember it about as well as you would, if your hopes for life were destroyed and then revived in the space of a week. 🙂 ) Before that, all I had was a Westinghouse project that didn’t win.

    But I think the issue goes deeper. Even if my paper had been accepted in time, it might have made some impression on faculty, but crucially, they’re not the ones who read undergrad applications! Admissions officers do, and both their own words and their “revealed preferences” suggest extremely different priorities from what the faculty might have.

    So even with the paper, as a 15-year-old who’d clashed with teachers at my high school and gotten some poor grades (including a C in chemistry), who’d left high school without graduating and gotten a G.E.D. instead, and who had no sports or music or foreign languages or “leadership,” I really don’t see how I could’ve demonstrated my readiness for Cornell and CMU (let alone for the schools that rejected me…) without the SAT or something like it.

  88. Scott Says:

    Art #86: One can define such a model in a stupid way, e.g.:

    “If n is composite, then an n-bit program halts immediately if its first bit is 0, or runs forever doing nothing if its first bit is 1. If n is prime, then an n-bit program is to be interpreted in your favorite Turing-universal language.”

  89. Julien Says:

    Hi Scott,

    Pardon me to be a bit corny. I’m usually a lurker. I am also pessimistic so I can’t cheer you. But as I understood, you and your wife are a bright couple and if I remember correctly you have a daughter. The knowledge that people like your family exist and that you pour efforts in defending science and the ambitions of nerdy kids that almost nobody cares about is really an inspiration.

    Take care.

  90. Mike Says:

    As anecdotal evidence for the whole SAT thing, I just took the SAT last year. Expensive prep courses don’t seem very useful—I know a lot of friends who took them, and they didn’t do much better than me. They also complained that they mostly teach you strategies that are essentially useless for anyone who already has decent test-taking ability. Checking out prep books from the library and doing free online practice tests works just fine.

    I have a high SAT score and barely any extracurriculars (I’d rather learn stuff on my own than participate in random inconvenient social rituals), so the new changes are really concerning for me. Hopefully they don’t spread too fast.

  91. Scott Says:

    Julien #89: Thanks so much!! (Besides our daughter, who’s 7, we have a son who’s 3.)

  92. Christian Says:

    I’m a undergrad student studying computer science, and I probably benefited from SAT/ACT test scores, though I ended up at a pretty uncompetitive school. I’m tentatively frustrated by the decision by UC, as I think the SAT/ACT offers a chance for any gifted student to turn 20 hours of study and a $20 book into an impressive score. In that sense, it’s much cheaper than tutoring or extracurriculars (or a “Side door” admission!). I think a prestigious system like UC, especially being public, should be very concerned with the financial arms race that goes into high end college admissions. I don’t know the proper ratios of GPA to Standardized test scores to arbitrary “holistic” metrics, but I think Test Scores should definitely be part of it. I also think the arbitrary “holistic” measures raise less ire, but create biases just as test scores do, only with less clarity and oversight.
    Sometimes I get frustrated at current news events, it seems that many good institutions keep eroding. The US government seems so much less effective than it used to be at protecting its citizens and responding to dangerous situations. That being said, I’m confident that things will improve and continue to improve, especially asymptotically. Trump, or out of touch lawmakers, or even uneducated voters, simply aren’t powerful enough to make any real change to the asymptotic behavior of the human race. While there are patterns of nationalism, nativism, and feckless leaders in the past, none of them yet have created any permanent, substantially negative changes to result in any past that I could prefer to the present. The world we live in now is vastly better than the 1800s, or even the 1950s. There may be an argument for n years in the past being better than current day for any specific country, but from a global context, we have better distribution of resources, and more access to information than ever before in the past. This makes it easier for a greater number of people to contribute to positive changes and new technology, and these trends are likely to continue. I’m optimistic about my future, and excited for what future generations may get to see.

  93. Kevin Zhou Says:

    One of the frustrating things about discussing the merits of standardized testing is that you always get people jumping in saying stuff like, “The tests are pathetically easy; I got a perfect score on them barely trying and then succeeded in life, so the tests serve no purpose.” When people make this complaint above, I can’t tell if they’re trying to humblebrag or if they legitimately think that’s a good argument. Of course talented people do unusually well on the tests and get rewarded for it — that shows precisely that the test is doing its job, not that it’s unnecessary!

    Anyway, speaking of politics, what are your thoughts on the newly proposed bill ( to ban all Chinese STEM grad students and postdocs?

  94. Scott Says:

    Kevin Zhou #93:

      what are your thoughts on the newly proposed bill to ban all Chinese STEM grad students and postdocs?

    Do you have to ask?? I’m horrified, ashamed for my country, and worried at a personal level for my Chinese students and postdocs. Hopefully this will go the way of the spite-driven proposal from a few years ago, to tax PhD students on their tuition waivers. But the Republicans will keep trying this shit.

  95. Raoul Ohio Says:

    John Cherniavsky #41

    The old “Why does it get hot in the summer” trick question is pretty bogus.

    As I recall, the effect of being close to the sun is about 1/3 of the effect of the tilt of the axis.
    At this time, the sun is closer during North winter, and South summer. The Southern hemisphere would have much more dramatic summer/winter asymmetry, except that the south has much less land, and water ameliorate the effect, except for central Australia, which is far from water.

    So making fun of people who guess “sun is closer” is lame, because they made a reasonable guess (much simpler than axial tilt), and are 1/3 right.

  96. Raoul Ohio Says:

    Chris B #52,

    Re the “bad old days” with ‘anti-communist’ lies”, …,

    This might surprise you, but actually there was some reason to be concerned about communism!

    For example, usual estimates put the number who died as communism took over Russia and China as 50 million each. If you don’t get 50 million, try counting to 1 million.

  97. barbara Says:

    Dear Scott,

    I dearly feel with you beeing depressed from the current overall situation with IMHO crazy leaders all over the world. Already some years ago, my father (born in 1946) pondered that in his youth, they fought for a better world (troubles back then: cold war, acid rain, still some old nazi-elite-networks here in Germany,…). some 40 years later, he was not sure, if their fight had been successful. I have never again raised this subject, as I too, feel the world becoming more and more insane.

    I feel comfort in some ideas brought up by other commentors like oscillating and self-healing societies on a macroscopic level, but feel sometimes lost on the personal level. Therefore I wanted to share the following quote which sometimes cheers me up:

    “We ourselves feel that what we are doing is just a drop in the ocean. But the ocean would be less because of that missing drop.” (Mother Teresa)

    Be sure, that your drop in the ocean is important not only to your family but to a lot of other people (including me). All the best!

  98. Aleksandar Totic Says:

    “So, what can anyone say that will make me less depressed?”

    I share your views of the state of America today. I grew up in Yugoslavia under socialism. I watched a great place to live tear itself apart in a decade. The downhill included toilet paper shortages, inane public discourse, “rapists on the border”, and general inability to do anything right. Incompetent leaders were making bad decisions, and spreading lies because they could not survive the truth.

    This is what I tell myself to cheer myself up:

    America is not in a stable bad state. There are highs, and there are lows, Obama/Trump, .com boom/financial crisis. And when something is not stable, there is hope for a quick turnaround for the better.

    And if it does not turn for the better, you have options worldwide. This particular ending does not cheer me up, but it does provide a bottom bound for my unhappiness.

  99. Anon Says:

    Churchill said that you can always count on Americans to do the right thing, after exhausting every other option.

    I am as pessimistic as you about the choices made, and think they terribly incompetent, but I am optimistic in that eventually after it becomes clear these options were bad, America will make the right choices in the end. Eventually this virus will be gone, whether from a vaccine or herd immunity or leaders finally choosing the correct decisions.

  100. jonathan Says:

    Sadly, in the span of just 8 years we’ve gone from two smart and decent candidates (Obama and Romney), to zero (I’ll give Biden decent).

  101. Anonymous again Says:

    I remember reading many years ago that IQ tests were bad at measuring intelligence because they measure very specific ways of connecting information, whereas other ways may be just as logical. I wonder if SAT and ACT testing is too dry, and lacks an extra dimension, to measure another quality – a burning curiosity to solve a real-life problem, which is boring to the mind of a particular type of person, for instance, that has been diagnosed with ADHD but gets so bored out of their mind by those tests that they don’t do well, yet are great at solving real life problems because they are not bored by those. Motivation is a huge factor for problem solving. I could stay up all night and do very well at solving a particular type of problem, but standardized test have always given me nausea. I do agree that the motive may be what you’re saying, though, and they may be better than the alternative. I don’t know, I don’t have the answers.

    It is difficult to exist in this world sometimes as someone who is intelligent and cares about the state of the world as a whole. I hope you feel better soon and at least find some solace in daily living.

  102. Gerard Says:

    @Raoul Ohio #95

    > So making fun of people who guess “sun is closer” is lame, because they made a reasonable guess (much simpler than axial tilt), and are 1/3 right.

    Your comment would be more poignant if we didn’t have universal public education, not to mention Wikipedia and Google.

    While I don’t think making fun of people is generally a productive activity I do think that this result is really quite appalling. It shows that many people have absolutely no curiosity about some of the most basic and significant physical mechanisms affecting our lives (unless, I suppose, you live near the equator). I’m skeptical that individuals with this characteristic could ever become highly effective in any productive activity that involves even a modicum of intellectual effort.

  103. fred Says:

    Things are relative.

    You may things are bad now, but for many people (inside and outside of America), things have been bad for a long long time. Yet, those people have endured, one way or another.

    As Phil Collins put it “Oh, think twice, ’cause it’s another day for you and me in paradise”.
    And when someone takes a shit in my paradise, I chin up and carry on.

  104. ryan williams Says:

    Thanks. I guess the point of my reply was to simply say that I don’t see the demise of the SAT in the UC system as evidence of the collapse of America. (Yes, there is other evidence… but I don’t mourn the SAT personally. I am skeptical it is such a great equalizer.)

    So, OK, here is another attempt to cheer you up.

    Recently I’ve been filling in a lot of gaps in my knowledge of quantum computing (I wanted some new intellectual distractions) and your various sets of notes have been a great resource for that. Thank you.

    What’s a problem in quantum complexity that my perspective might help with?

  105. anon Says:

    How much decision-making power do you have at UT Austin?

  106. Scott Says:

    ryan #104:

      What’s a problem in quantum complexity that my perspective might help with?

    So glad you asked! 🙂

    Is there a subquadratic quantum algorithm for edit distance? Or can we give complexity-theoretic evidence against it? (Recent results weakly indicate that n3/2 might be possible.)

    Is there a better-than-2n classical algorithm to spoof quantum supremacy experiments like Google’s? (That is: given a large-depth, random quantum circuit C on n qubits, to sample output strings whose probabilities of being output by C are appreciably larger than the average of 2-n?)

    Or what about an AM protocol, or even a QCAM protocol (that is, classical Merlin and quantum Arthur), to recognize such strings, in less than ~2n/2 time (to allow for the Grover speedup)? Such a protocol, if it existed, would break the security reduction for my certified randomness scheme—i.e., for arguably the only known application for a QC that’s already been built. 🙂

    In my Oracles Are Subtle But Not Malicious paper, I claimed to show that if PP is in BQP/qpoly then the counting hierarchy collapses. But my 1-paragraph “proof” had an error. Does such an implication hold?

    Is there an exponential lower bound on multilinear formula size for an explicit function f:{0,1}n→R? Would such a lower bound need to be non-natural? Is there an efficient algorithm to learn small multilinear formulas? For the applications to QC, see my 2003 Multilinear Formulas and Skepticism of Quantum Computing paper. Section 10 of that paper also gives a lower bound technique for a certain kind of multilinear formula that totally looks like it evades the natural proofs barrier, by exploiting symmetries in specific functions (but again, I don’t know whether a natural proofs barrier existed there in the first place).

    Please let me know if you solve these and need more! 😀

  107. Scott Says:

    anon #105:

      How much decision-making power do you have at UT Austin?

    Extremely little—UT is a huge institution, with over 50,000 students and 3,000 faculty. Even within the CS department, I’m only one vote among dozens.

  108. Chris Blanchard Says:

    @Raoul Ohio #96

    I quite agree with you. There was an awful lot to be concerned about from communism, with Russian conquests running murderously amok, but McCarthyism and the like seriously weakened the US response to those conquests. In particular, Truman’s administration destroyed the Office of Strategic Services, which had been very effective, and replaced it with the CIA, which was not. The OSS ran successful operations in war time Germany, Turkey, and all over the place. It wasn’t replaced in one lump but its immediate precursor was in place in good time for it to fail to spot the Soviet conquest of Czechoslovakia, and the full thing was in place in time to not see the Chinese invasion of Korea, with 300 000 troops massed for conquest. Not a good record then, and followed by, amongst many other failures, their take-over of South Vietnam in 1955. This change (OSS to CIA) had several causes, as these things do, but part of it was designed to get rid of any kind of left wing influence, which they called communism, or fellow travelling, or being a ‘pinko’, so they sacked most of the people with a real record of war time intelligence work and replaced them with ‘anti-reds’, many of whom were inadequate, and who were rapidly infiltrated by numbers of real Russian spies. There were Russian spies in the US, but McCarthyism made their lives easier. Russian Intelligence wasn’t stupid, so my best guess, and we know too little for it to be more than a guess, is that they encouraged the divisive and destructive right wing, to damage the USA. We know that the successors to the Soviets are doing exactly that now, so what’s different?

  109. Gerard Says:

    Scott #106

    > Is there a subquadratic quantum algorithm for edit distance? Or can we give complexity-theoretic evidence against it?

    That later sounds like it would be quite a challenge. Unless I’m mistaken no one has ever proved a superlinear lower bound even for 3-SAT.

  110. Scott Says:

    Gerard #109: I didn’t say “unconditional proof.” The goal would be to show, e.g., that if there’s a subquadratic quantum algorithm for edit distance then there’s also an O(1.99n/2) quantum algorithm for k-SAT, analogous to what’s already been done in the classical case.

  111. ryan williams Says:

    Scott #106:

    Thanks! I strongly suspect what is called “IQP” (and related low-depth stuff) admits a generic less-than-2^n classical speedup, for most formulations that I have seen. I haven’t gotten around to carefully reading Google’s particular setup, but “high depth” would probably trip up my ideas (which revolve around less-than-2^n algorithms for counting Boolean solutions to low-degree systems of polynomial equations). Maybe throwing in Merlin would help.

    I’ll think about the others…

  112. Scott Says:

    ryan williams #111: For very low depth, 1D or 2D spatially local, random quantum circuits, Napp, La Placa, Dalzell, Brandao, and Harrow have already given heuristic evidence that a classical spoofing algorithm exists.

    Meanwhile, for arbitrary random quantum circuits of low enough depth, Barak, Chou, and Gao recently showed that fast classical spoofing is possible, albeit with an advantage that drops exponentially with the depth.

  113. Gerard Says:


    A little while ago I suggested to you that the notion of oracle/black-box problems was likely to be very confusing to non-specialists.

    Well here’s an article in Forbes that I would argue manages the feat of getting almost everything it says wrong and I think a good part of that can be traced back to the author misunderstanding an oracle separation result:

  114. Scott Says:

    Gerard #113: No, I disagree. I didn’t see anything outright wrong in that article (in fact it gets several things right that most articles don’t), but it certainly misleads readers by prominently leading with the Traveling Salesman Problem — only admitting near the end that that’s not one of the problems for which we think there’s a big quantum speedup! But that misleadingness has nothing to do with misunderstanding Raz-Tal or any other oracle result. Ironically, in fact, the misleadingness could’ve been prevented had the author understood and internalized a different oracle result (the BBBV Theorem)!

  115. Jacob Says:

    A more optimistic take on how the US is doing OK despite Trump’s idiocy. In the end the death rate is similar to the EU. Slightly better than the Netherlands which I wouldn’t call a collapsing Leviathan.

  116. Gerard Says:

    Scott #114

    So how do you explain this passage ?

    > Fortunately, many computationally difficult problems — including, quite probably, the travelling salesman problem — are far less difficult (and far less computationally expensive) using a quantum computer. It was proven, just a few years ago, that quantum computers possess a computational advantage over anything a classical computer could ever achieve.

    Note that the proof he referred to here is this oracle separation result:

    The only way I can see this statement as likely being correct is if he is talking about the quadratic speedup of Grover’s algorithm. But I doubt very much he’s merely referring to a quadratic speedup and in any case if he is talking about Grover I don’t see what that has to do with an oracle separation between BQP and PH, so why mention it here.

  117. Scott Says:

    Gerard #116: You’re right; I’d missed that sentence, which is bad! But in some sense, the main problem with it is not that Raz-Tal (which it’s referring to) is an oracle result. Rather, the main problems are

    (1) this has nothing to do with the Traveling Salesman Problem, and

    (2) the cited statement (in the oracle setting) goes all the way back to Bernstein-Vazirani in 1993 and Simon in 1994. Raz-Tal solved the much harder problem of relativized BQP vs. PH.

  118. Gerard Says:

    Scott #117

    I mean there are so many issues with this passage that it’s hard to know where to begin. Let’s consider (to simplify things greatly) two possibilities for the world in which we live:

    1) The conventional view: P != NP, BQP not in P, NP not in BQP.

    2) The alternative view: P = NP = BQP = P^#P.

    Well the author’s claim that TSP is much easier for a quantum computer compared to a classical computer is false in both of these worlds.

    Moreover when he says “It was proven, just a few years ago, that quantum computers possess a computational advantage over anything a classical computer could ever achieve.” he leaves out the key qualifier: “relative to an oracle” (which admittedly very few of his readers would understand anyway) .

    Most people, myself included (if I didn’t already know that it was false) would believe that this statement means that it has been proven as a mathematical fact (like 1 + 1 = 2) that there are problems for which a quantum computer is superior to any possible classical computer.

    Now you say that all this actually has nothing to do with the statement he cites but the fact that he cites it is evidence that in his mind he somehow came to believe that the BQP vs PH oracle separation somehow implies his two highly inaccurate conclusions. Which I think supports my point that if computer scientists talk about oracle results to non-specialists, let alone the general public, they really should emphasize that these results are purely theoretical constructs which are not known to have any direct relevance to the real world.

  119. Scott Says:

    Gerard #118:

      these results are purely theoretical constructs which are not known to have any direct relevance to the real world.

    No, that’s too strong of a statement—or rather, it’s true only in the sense that even most circuit complexity results also don’t have “direct” relevance to the real world. Computational complexity and oracle/query complexity are simply two different theoretical models of computation. You’re right that if we care about, e.g., whether NP-complete problems can be solved in polynomial time, then computational complexity is much more directly relevant, with query complexity only an indirect indicator (albeit, one that’s already good enough to refute at least half of the misconceptions about quantum algorithms that actually occur in practice…). If, on the other hand, you care about super-efficient (say, logarithmic-time) algorithms for querying extremely large inputs, as in “Big Data,” then query complexity is just the model you want.

  120. Gerard Says:

    Scott #119

    There are aspects of this that I’m still confused about, especially with the way black boxes are used in quantum complexity theory vs. classical complexity theory.

    In the classical case my objection to them is that in the real world there’s simply no such thing as a black-box. If you’re given some way to compute a function then you always have the ability to look inside it. Also sure you can say that you only care about how many times you have to call the oracle and not about how long the oracle takes to return a result, but that’s also a made up restriction that has no practical relevance. In practice the only thing you are going to care about is how long the computation actually takes. For those reasons classical black box results seem completely artificial to me.

    I’m less certain about quantum black box results, especially ones that don’t involve exponential speedups. For example Grover’s algorithm is stated as a black box result and until very recently I thought it would continue to give you a quadratic speedup even if it turned out that P = P^#P. Now I’m not sure about that. I’ve also seen it stated as involving database queries but then the question arises how did you input the database ? It seems like the input process is going to be order N, so is the speedup even real ?

    I suppose that when you get into lower complexity classes the details of the computational model start to matter a lot more than when you only care about exponential speedups.

  121. Scott Says:

    Gerard #120: Let me try again. The black box model closely models reality in situations where
    (1) you have random access to a large database, and
    (2) queries to that database (rather than other computations) are the limiting resource.
    This is true both classically and quantumly. In the quantum case, however (e.g. with Grover’s algorithm), assumption (1) would be satisfied only if one built a “quantum RAM” (qRAM), which is another technology (besides scalable QCs themselves) that seems possible in principle although it hasn’t yet been demonstrated.

  122. Gerard Says:

    Scott #121

    OK, but I’ve never heard of a computer that came with a large DB pre-loaded in RAM. I guess I can’t prove that there’s no practical application of this, but I’m having trouble thinking of one.

  123. matt Says:

    Scott, for this question “Is there a better-than-2n classical algorithm to spoof quantum supremacy experiments like Google’s? (That is: given a large-depth, random quantum circuit C on n qubits, to sample output strings whose probabilities of being output by C are appreciably larger than the average of 2-n?)”, can you make the point at which results become “interesting” more precise?

    For example, if the time is 1.99^n and the outputs the classical algorithm gives have, on average, a 1.99^{-n} probability of being chosen by the quantum algorithm, is that interesting? What if that algorithm uses space 1.99^n also, is it still interesting?

    And what kind of spacetime circuit geometry are you considering? If you have a 2 dimensional circuit of size L-by-L with n=L^2, and depth d, then there is a 2^{dL} algorithm. So, are you keeping d fixed but going outside 2 dimensions to some expander geometry for the ciruit? Or are you imagining that you scale depth proportional to sqrt{n}? Again, where would that be interesting?

  124. Scott Says:

    Gerard #122: There are plenty of situations where you’re maintaining a lot of data in RAM, and you want both the updates and the queries to be fast (say, log(n) time). Indeed, the whole field of data structures is basically about that! If you like, query complexity is the simplification where we treat the data structure as given and focus only on minimizing the query cost.

  125. Scott Says:

    matt #123: My answer to all of your questions is the same—I personally think any result in this direction is interesting, to whatever extent it nontrivially beats what we already knew. As examples, I thought the Napp et al. and Barak et al. papers were both great, even though the depth in both cases was much too small to challenge the Google experiment.

  126. mjgeddes Says:

    Good news Scott, after a period of intense reflection on quantum mechanics, I think I may have an insight into what the correct interpretation is !

    I have to clarify my thoughts a lot more, but at this point I’m happy to stick my neck out and state that the Many-Worlds Interpretation is false!

    So basically, I realized that there was an analogy between wave functions and language ; basically I think wave functions are analogous to data models , programming languages and mathematical logic. So QM is a language we use to reason *about* reality, wave functions do *not* refer to objective reality. Wave functions, I think, are no different to boolean logic or the python programming language, they are human inventions.

    So what is the underlying reality? Basically, I think it’s geometry, not ordinary space-time, but something more fundamental. So it’s a hidden variable theory, but not like any of the current proposals. The wave-functions are the *language* used to talk about this new geometry.

    OK, I just want to point to a recent article in Quanta magazine, which I think is very relevant to quantum foundations, entropy is linked to energy:

    I think that there’s a new interpretation of QM on it’s way, and it’s hopefully going to blow the others out of the water, and finally clear up quantum foundations.

  127. Vick Says:

    As a Canadian, the only reason I ever took the SAT was on the off chance I’d apply to US universities. And the whole thing was a scam. I was signed up to special courses, hosted by high-scoring SAT alumni, and taught the exact way to write an essay in a way that appeals to SAT scorers. I ended up getting near perfect on reading and writing, but lost a few points in math on trivial problems because of clumsiness/bad luck and the difficulty to demonstrate actual math skill in multiple choice, but my final total score of 2320 still didn’t get me any meaningful attention from the US schools I applied to. So yeah, bullshit.

  128. Gerard Says:

    Scott #124

    > If you like, query complexity is the simplification where we treat the data structure as given and focus only on minimizing the query cost.

    In my experience significant algorithmic improvements usually involve either using better adapted data structures or some more fundamental change in problem representation, so I’m not sure how meaningful lower bounds that exclude such changes would be. That also further narrows the possible utility of Grover’s algorithm. If the number of queries you need to perform on the same data is significant it makes sense to pay the n * log(n) sorting cost and then you have a query cost of log(n), which is way better than Grover’s sqrt(n).

  129. Scott Says:

    Gerard #128:

      If the number of queries you need to perform on the same data is significant it makes sense to pay the n * log(n) sorting cost and then you have a query cost of log(n), which is way better than Grover’s sqrt(n).

    OK, and what if it’s a database with lots of attributes, and you need to support queries on complicated contributions of the attributes? Then sorting won’t work, and Grover might be the best option.

    Having said that, I agree that the application of Grover to combinatorial search problems has always seemed the most compelling to me, since it doesn’t depend on assumptions about a qRAM and how easy it is to query and why its contents couldn’t have been organized differently.

  130. Gerard Says:

    Scott #129

    > OK, and what if it’s a database with lots of attributes, and you need to support queries on complicated contributions of the attributes?

    But does the proof of O(n) optimality for the classical case apply then ? At least 60 years of work have gone into making those kinds of things efficient, which I think can be more or less summed up with one word: indexes.

    > Having said that, I agree that the application of Grover to combinatorial search problems has always seemed the most compelling to me

    Right, but what I wonder about is what, if any, quantum speedups would survive if P = NP and/or P = P^#P. So far Grover in the database search context seems like the only possible candidate.

  131. Gerard Says:

    To clarify my last comment, I meant that Grover is the only candidate if P = P^#P. If P = NP but #P is not in P then there might be more possibilities (like quantum physics simulations).

  132. Sniffnoy Says:

    Vick #127: As has been mentioned upthread, when this has actually been studied on any large scale, the results have suggested that in fact such classes barely raise scores. Rather than being bullshit, the test in fact appears to be pretty robust against such things. And, as Scott and I have mentioned upthread, the relevant question isn’t “is it bullshit”, but “is it less bullshit than what will be used instead” (which seems pretty plainly true). Anyway, see my longer comment upthread for more.

  133. Anon Says:

    You cannot have a smart person in charge if most people are dumb.

  134. alex Says:

    My optimism comes from the blog post you mentioned writing sometime in response to Wolfram’s latest nonsense. I enjoyed your old review of NKS and I am looking forward to reading what you have to say

  135. Doug K Says:

    as an antidote to despair I find Teri Kanefield helpful, see terikanefield-blog dot com. Posts tagged ‘solutions’ in particular.

    For myself I try not to think too much. This is by Jorge Luis Borges, read it first at age 17 and it’s helped ever since to survive episodes of contemplating quietus. There are always ‘reasons more terrible than tigers/to prove to us that wretchedness is our duty’

    a man who has learned to express thanks
    for the days’ modest alms:
    sleep, routine, the taste of water,
    an unsuspected etymology,
    a Latin or Saxon verse,
    a man who is aware that the present
    is both future and oblivion,
    a man who has betrayed
    and has been betrayed,
    may feel suddenly, when crossing the street,
    a mysterious happiness
    not coming from the side of hope
    but from an ancient innocence,
    from his own root or from some diffuse god.

    He knows better than to look at it closely,
    for there are reasons more terrible than tigers
    which will prove to him
    that wretchedness is his duty,
    but he accepts humbly
    this felicity, this glimmer.

  136. Leonard Susskind Says:

    Dear Scott,
    I, like you, was a bad student and I didn’t “have a paper.” I’d been asleep at the wheel for twelve years. There was no way that I could have gotten into college without standardized testing. But you and I are a tiny minority, and frankly the system did not fail us in the end.

    However, there is a huge number of young people out there (some of them just as smart as you and surely smarter than I) that our American system is totally failing. I don’t really believe that the source of the failure is the SAT exam, but the fact remains that the problem of maverick Jewish misfits is a very minor problem in our society, when compared with the problems of young black and brown kids who have no chance at all. Have some perspective.

  137. JimV Says:

    Way back when I took the SAT (circa 1964) it seemed like a pretty good pair of tests (verbal and math) to me. Comparing my scores to those of smart people I met in college seemed to indicate it was a good predictor of “smartness”. I went to a small rural high school (graduating class of 19 students), had no SAT preparation, and got accepted at several universities largely on the strength of test scores. I was a very un-well-rounded introvert. (I got better, as they say in Monty Python’s “The Holy Grail”). A lot of the people I met in college weren’t learning much there (nice guys though, mostly). Those were the ones who didn’t have great SAT scores. I don’t think they would have been helped much by SAT prep, speaking as the guy who tried to help them with their course work.

    Much later in life, I won some engineering awards at GE and had an upper-level technical manager tell me, “Every problem we give you, you solve.” (By engineering standards, nothing earth-shaking.)

    Anecdotes don’t mean much and maybe the SAT has gotten worse over time, but I will certainly be sad to see it go. You need a standard to judge people, fairly, across different backgrounds and personalities. If the test is bad, improve the test.

    I wish I could cheer you up. Well I’ll try the old standard I used on my nephews and nieces when they seemed depressed: “You have a family who loves you.” (And a bunch of faithful readers.)

    About Nixon and Reagan: they had mostly Democrat-controlled congresses. If Nixon had had Trump’s Senate, look out!

  138. Depressive Realism Says:

  139. Raoul Ohio Says:

    Chris B #108,

    Pretty interesting. You have a deeper understanding than I realized. I will follow my own frequent advice: “Don’t argue with people who know more than you do”.

  140. Richard Gaylord Says:

    you wrote “Look, I’m a theoretical computer scientist. ”

    can you comment on Feynman’s comment

    note: i am not saying that there’s anything wrong with engineering, to quote Seinfeld

    i went to the Polytechnic Institute of Brooklyn which is (actually, was since it no longer exists) a premiere university (we used to say that it was better than MIT because only New Yorkers went there) but i never took a single engineering course; just chemistry, physics, math).

    i’m asking, not just to give you something less depressing to think about than the present and the future of this country) but because it seems to me that the term ‘social science’ is used by that community in an attempt to piggyback on the street cred that theoretical physics has with laypersons (not because of the subject per se but because of the creation of the atomic bomb and because of Albert Einstein).

    note: i am also disturbed that none of the pop science magazines seems to think that chemistry is a science (at least there is almost no coverage of chemistry in quanta, etc). and the subject of philosophy of science is really philosophy of physics which is entirely different from chemistry (note: in addition to my B.Sc. in chemistry, my father got his Ph.D. in chemistry – from Brooklyn Polytech as it happens).

  141. Scott Says:

    Leonard Susskind #136: It’s an honor to have you here as always!

    I was fascinated to learn that, even though you hold a different opinion, your actual experience was so similar to mine (despite being separated by 40 years).

    No, the system didn’t fail you or me, although at least in my case (I don’t know enough about yours), I’ll give the system credit for a pretty good try. My worry is precisely that the system is now set up to fail the next Lenny Susskind—who’s as likely to be Chinese or Indian or whatever else as to be a Jewish kid from New York.

    And I don’t think our system serves black and brown kids terribly well either—not least, because it seems primed for regular right-wing backlashes (Reagan, then W, and now the worst one of all). I’ve become a strong supporter of a Universal Basic Income, which would establish a baseline of hope for everyone in the country. I also support affirmative action, even in the form of explicit quotas (though I think looking at economic disadvantage rather than directly at race is another idea worth exploring). Most importantly, I support the total dismantling of preferences for legacies, for children of wealthy donors, and for the “patrician hobbies” (horseback riding etc.)—i.e., the regressive set-asides that together account for a huge fraction of the incoming class at schools like Harvard and Princeton.

    I actually agree that, even if you tie the right hands of sufficiently smart kids behind their backs (as happened in the US, the USSR, and many other countries throughout history), many will still succeed with their left. In that sense, our system can survive a lot. What it can’t survive, though, is a complete rejection of the concept of intellectual merit.

  142. Gerard Says:

    I would say the whole way meritocracy is implemented needs to be rethought from the ground up.

    Implement a UBI sufficient to cover everyone’s basic needs but don’t put any bounds on what people can accomplish if they have the will and the ability.

    Even if access to elite education were more fairly apportioned, as it probably is in a country like France, where there is a well established system of entrance exams for the most elite institutions and a different system (still exam based) for the Universities as a whole, I still think the system would be unjust and suboptimal because it puts all the weight on what a person does between the ages of say 14 and at most 30, with most of the weight falling in the earlier to mid part of the range. But people change a lot over time and, whether for genetic or environmental reasons, not everyone develops at the same rate. For example despite doing very well in high school I struggled in college because I hadn’t yet developed sufficient intellectual discipline to deal with the much more challenging work (both in terms of difficulty and quantity). In addition to that I also had significant emotional issues during that phase of my life. It took me several more decades to fully overcome those issues (if indeed I have even now) and by then my life’s course was largely determined.

    It seems to me that today the whole system of barriers created by the educational system is largely unnecessary. Access to information at all levels is practically universal (a process that should be completed by the elimination of paid journals), in many fields significant contributions can be made with little more equipment than a laptop, and here I am, just some mostly anonymous guy on the Internet, having conversations with one of the top TCS professors in the country.

  143. Arboster Kipling Says:

    This is what happens when you lock yourself inside of a bubble. Scott you DID make a prediction of sorts when you said 250K deaths was the best case scenario. Given that we don’t really have great data and that we are likely over-counting being at 100K is sad but not unfathomably terrible. Its not over yet but we’ve a while to go before we get to your ‘best’ case. I imagine you might prefer the kind of governance in Sweden where there are more deaths per capita than we have. The same goes for France, the UK and many other Western democracies not ruled over by Cheeto Hitler. As others have pointed out the per capita deaths in our peer Canada are not staggeringly different, if they were our size there’d be around 70K deaths) and its clear many of the US deaths are the result of Democrat governors forcing nursing homes to take covid patients including my governor who got is PhD at an obscure little school called MIT. The 1957 flu killed 116K people when America was half the size and there wasn’t a huge shutdown or people claiming it would be the end of the world. Obama raided the national stockpile of PPE and did not replenish it during his 8 years. Yes Trump should have done it in his first 3 but I won’t argue Trump is a paragon of competence. Just relax and thanks for the lecture notes from your last post.

  144. Christopher Blanchard Says:

    Raoul Ohio #139.

    Thankyou. That is very kind, but please … . I am just a dilletante and lots of much better informed people will certainly disagree with me. I am here because I got into my sixties and decided a decently cultured person, which is what I try for, should understand quantum mechanics. I managed Roger Penrose’s “The Road to “Reality” and I have read Scott’s Complexity classes (P/NP), but this stuff is dizzying, so all I can do is throw in a half relevant bit of what I do half know.
    For a bit of optimism, and relevant to the debate about SATS. I got chucked out of ‘A’ level maths, basically for being idle (that is UK maths for 16 to 18 year olds), have a first degree in the silly pseudo-science called economics and an MBA – which is also pretty silly but, and this is the optimistic, I can still, with help from the kind of people who write here, learn about material on the world’s leading intellectual edge, even in my sixties. Yo!

  145. Yiftach Says:

    Music is a great way to improve the mood, for example, Ian Dury’s Reasons to be cheerful, part 3

    No one seemed to mention it but it is important to recall that depression is a serious medical condition, with emphasize on medical. You might like to take a medical advice rather than advice from random people on the internet. (You are smart enough to rationalized your depression, but it does not mean it is not a medical condition.)

  146. Anon 4 Says:

    Agree with you on the SAT. Not sure if this contributed to the difference between the moon landing NASA of yesteryear and the botched testing rollout a few months ago but the Carter administration ditched standardized civil service hiring exams in the name of “diversity.” These have been used in China for over a thousand years to ensure competent bureaucracy. I see no hope of this being fixed anytime soon as the current zeitgeist (and even very high IQ readers of your blog) has a major blind spot to any tradeoffs between diversity versus competence, objective fairness, and cohesion.

    Less importantly, if those at home COVID tests start become available I hope they work better than the Gates Foundation’s IHME projection model. For Gates genuflecting readers who romanticize his status as a billionaire non-state actor like Bruce Wayne or an Ayn Rand protagonist, I strongly urge you to instead consider risk in your geographic area based on the model at since it has a much better track record.

  147. Filip Says:

    I’m writing this half an hour after the launch of Crew Dragon: I’m glad this happened, but isn’t it kinda sad too? I’m a progressive liberal but I always felt that the most brilliant Russian/American engineers truly served their countries during the Space Race. Now we need an Twitter troll billionaire (that didn’t believe in covid exponential growth until very recently) to do the same.

    I don’t want this to be a hot take, so I’m interested in other people’s opinion — is there a bright capitalist future of spaceflight? What caused it, budget cuts? What can we expect in regards to quantum computing as well?

  148. Scott Says:

    Arboster Kipling #143: We’re on track for >250,000 US deaths by the end of the year just assuming the trends since April continue (say, ~1000 deaths/day with see-sawing falls and rises), and there’s not a huge new exponential spike as we relax the lockdown. And yes, if we’re super lucky we might even come in under that. I agree that we avoided the worst-case scenario, probably because of the lockdown that states and communities embarked on despite the void at the top.

    Incidentally, the spike in all-cause mortality suggests well over 100,000 deaths by now. And Canada, which did better than us, would probably have done better still if (like Australia and New Zealand) it didn’t have the misfortune of being next to us.

  149. Deepa Says:

    The more subjective system is, the more it encourages (otherwise nice) parents and kids to become borderline sociopaths (at least in the high school years), and rewards them for those behaviors. 

    I don’t mind a subjective system for admissions as long as an expert is assessing your application. For this, they would need to make you commit to a major and apply only to that. And an expert from that department would process applications. Even an interview would make sense then.

    I support harder tests. For example, if you apply for a math-intensive major, you ought to have a certain score in the math olympiad. I wonder why some universities don’t start doing that. Wouldn’t they attract candidates who will prosper wildly at a hard program? And do great things later? In this way they could attract away the best students from competing colleges.  This is what I don’t understand – you’d think the marketplace would create colleges that do this. Example : The IITs in India don’t care about anything but your score in 3 hard tests they design – in math, physics, chemistry. They have a trivial test for English now, I think. This type of system creates other problems, but it cannot be gamed by sociopaths.

  150. Gerard Says:

    Deepa #149

    > For example, if you apply for a math-intensive major, you ought to have a certain score in the math olympiad.

    Unless things have changed a lot in the last 30 years (which I admit is entirely possible) I think that would further increase geographical disparities. Growing up in a medium sized city in the Northeast I never heard of the math olympiad, and I certainly didn’t know anyone who’d competed in it (though there was one guy a year or two ahead of me who ended up going into a a PhD program in string theory at Princeton). When I got to Cornell I found that despite having been near the top of my class at what was considered one of the best high schools in my area, people from places like NYC and Westchester or other major cities were often much better prepared in math and science.

    I know an approach like that is used in a few fields, music for example. But it seems to me that science and math education at the secondary level is far too underdeveloped in most of the country for it to be generalized.

  151. Anonymous Ocelot Says:

    Filip #147

    You need to learn more about SpaceX 🙂 I think you’ll have a different viewpoint.

    The thousands of incredible spacex engineers have worked incredibly hard toward this day (and will continue working hard so that you and I can enjoy a multiplanetary future!).

    I don’t know why NASA/JPL slowed down; budget cuts seem like a good guess. But the ingenuity and innovation displayed by SpaceX has brought launch costs down by half so far (you can check wikipedia and compare Falcon 9 payload to LTO / $ vs Soyuz 2 and Atlas V) and will continue to push us further, beyond what Apollo was able to do.

    This is driven by a shared passion: SpaceX employees (much like their chief engineer) are obsessed with space, and in particular with humans reaching Mars!

    You are complaining about the way this goal was achieved; but there is nothing “sad” about these hardworking people achieving what they set out to do, and in fact it’s one of the few illustrations that human engineering hasn’t stagnated since the 60s. I hope you will not see it as “sad” anymore by the time SpaceX is taking us back to the Moon and, eventually, to Mars!

  152. Filip Says:

    Anonymous Ocelot #151
    Maybe I came off too cynical, but:
    (1) I totally agree that it was an amazing feat. Also tips the reality a bit positively while other mess is happening around.
    (2) I love what SpaceX engineers did. I’m aware of the launch cost optimization.
    (3) I guess I care about the engineers too much. They would’ve been national heroes and now are merely employees of billionaire space race companies. Take a look at the Wiki link and you’ll see why “passion” is not the motivation of such organisations — in fact they are cruel to each other!

  153. mjgeddes Says:

    What on Earth is happening over there in the States ? Looks like 60s all over again – manned rockets launching, rioting in the streets, the pandemic etc. !

    Not sure it’s really a collapsing leviathan, but rather merely a continuation of what has always been happening since the beginning of the industrial revolution – the history of the US has always been fairly tumultuous, and you’re just coming out of an unusually quiet decade or two and resuming the usual bedlam.

    In the end , there’s only two stable attractors, human extinction or friendly super-intelligence, and one of them will be reached within the next 2-200 years. This is, in the words of Hanson, ‘the dream-time’, a short period of extreme disequilibrium that started with
    the industrial revolution, and must end with either extinction and/or superintelligence establishing a new equilibrium. A period that’s but the blink of an eye in evolutionary terms.

  154. Deepa Says:

    Gerard #150,
    If it was the official criterion to make it into a particular set of majors in college, then of course, you’d have heard about it. There would (I imagine) be free training material available to everyone online.

    It would make the math section of the SAT unnecessary.

    I like one hard entrance test (like the math olympiad) to college. The SAT (or ACT) doesn’t distinguish between a lot of students. Even the subject tests do not.

    It doesn’t have to be the math olympiad exactly. It could be the computing olympiad or something else of that level of difficulty. Problems you haven’t seen before that require creativity. And nothing else should matter, not even grades in high school.

  155. Ashley Lopez Says:

    Deepa #149 (and others)

    But wouldn’t the math olympiads and ‘IIT-JEE’ kind of tests (that is the entrance exams for the IITs in India) filter out, for example, a future Albert Einstein? For some reason I cannot visualize Einstein scoring in a math olympiad.

    I am not familiar with the SAT tests and they could be a different story.

  156. RaceSpace Says:

    @Filip #152

    What horrors did the wiki contain? Sure, companies competing for the same market will compete. Two of those had a few lawsuits and – gasp – a twitter fight. That’s significantly less vicious than a thousand different clashes between high-profile companies, and even less vicious than a fight I’ve heard of between two companies producing cribs easily attachable to a mother’s bed.

    As for motivation – yeah, companies want money. Is that so much worse than “show up the commies”? Or even “engage the commies in a resource-heavy race so that they get hurt worse than we do”? Personally, I’d take companies and greed over this kind of patriotism and “national heroism”. And of course none of it reflects the motivation or the well-being of the engineers themselves.

  157. Scott Says:

    Ashley Lopez #155: Contrary to myth, Einstein was an excellent student in math and physics (a little less so in other subjects). Of course you’d never know from test scores that he’d be Einstein, but he would’ve done fine under Deepa’s system. While improvements are possible, I do like Deepa’s system better than what we have now.

  158. Gerard Says:

    Scott #157

    > Contrary to myth, Einstein was an excellent student in math and physics (a little less so in other subjects).

    I don’t think that’s entirely accurate, or at least needs qualification. He may have done well at the secondary level but he finished his undergrad Physics degree second to last (the last placed student didn’t receive her degree) and that was at a lower rated school in a tiny country (it only became a famous and elite institution after, and I think, because Einstein attended it). Under today’s system in the US I don’t think he would have gotten into graduate school, or at least not into a good one.

  159. Gerard Says:

    Another great mathematician who might not have done well under that system is Grothendieck. He seems to have had trouble learning things from others and needed to figure everything out on his own more or less from scratch. I don’t really know anything about the format of the math olympiad but I suspect that even a great genius wouldn’t have time to solve all the problems from first principles.

  160. Anonymous Ocelot Says:

    Filip #152

    I don’t care about a “billionaire space race”. I care about SpaceX, the company of passion who is pushing our abilities forward against tall odds. The engineers are absolutely heroes, and that is what I hear from the press (it’s been a lot of NASA and SpaceX patting each other’s people on the back … and virtually zero coverage of SpaceX vs other billionaire companies. Only thing I saw was some mention of SpaceX vs Boeing, which is different).

    That article was fine. SpaceX sued and won to get a patent nullified … something that makes me smile every time it happens in the private sector. There was sparring over PR events because these companies are competing for funding, and there was legitimate disagreement about the terms used to describe certain milestones. I really don’t see anything wrong, here.

    My point is:

    If you actually care about those engineers, then please reflect on whether they would appreciate you telling everyone how their incredible achievement Saturday is a “sad” thing in some really complicated, subtle way that has nothing to do with Space or engineering.

    (Btw … the engineers will make LOTS of money when SpaceX goes public; another wonderful feature of this being a private company (that uses equity effectively)! If you’re pining for, say, a nationalization of SpaceX, then you’re hoping for shrinking that payout to a fraction of what it could be otherwise … something the engineers CERTAINLY wouldn’t thank you for! And btw stock buybacks are a thing, so they can get some fractional payout now … it’s not IPO or bust. But, I’m sure most of them hold onto a lot of stock, and know that that stock will be worth a LOT more once their goals are achieved. Starlink will probably be the first phase of that, then Mars).

  161. Gerard Says:

    There’s a question I’ve wondered about all my life and this seems like as good a place/time to ask it as any.

    What is mathematical intelligence and why/how does it differ from verbal intelligence ?

    There seem to be 3 types of people who do well on the SAT (at least the old style SAT of 30 years ago or so), those who have high math and verbal scores, those who score higher on math than verbal and those who score higher on verbal than math (I was in the 3rd category and despite being a Physics major I probably had a lower math SAT score than the average Cornell engineering student – I guess I’ve never been one to take the path of least resistance in anything).

    But what is it that distinguishes mathematical reasoning from verbal reasoning ? It seems like much of mathematics involves reading, understanding and remembering theorems and proofs, which are really just sequences of logical statements, so I have trouble seeing how that is distinct from verbal intelligence. Of course there’s also geometric intuition but I have a hard time seeing how that helps in solving sequence completion problems, which, as I recall, were a large part of the old SAT, or doing complicated integrals.

    Is there some kind of fundamental pattern matching function in the brain that applies to both numbers and equations/formulas but not to normal language and that people possess to varying degrees ?

  162. rrg Says:

    I’ve been thinking about moving to a country where meritocracy is stronger. Maybe Taiwan, since I’ve been learning Chinese from my labmates and they’re a democracy. Their covid19 response has me thinking about it more seriously. I’ve been to Taiwan a few times but I don’t know much about them really; I’m worried I’m engaging in some grass is always greener wishful thinking.

  163. anon Says:

    For gifted mathematicians you are never going to get an honest response because they rarely open up about their life experiences and how differently they perceive the world (and they are smart enough to hide it well!). Plus, unlike athletes where skill is apparent to everyone, math is subtle and social scientists cannot objectively assess it.

    You can read Terry Tao’s blog on intuition vs rigour, and I think we can all agree that verbal intelligence is very important to communicate intricate ideas; but is it enough? Read Euler’s, Newton’s, Einstein’s, Erdos’, Grisha’s, Nash’s biographies and you will see severe obsession with details and inability to perform tasks that are trivial to the average human. It’s obvious there is some divergence in mental functions, and there is a neurodivergent pattern with their parents, siblings and children.

    Jews represent a quarter of Turing and Nobel laureates despite being less than one percent of the world population. Kenyans are extremely overrepresented in distance running. And so on. Our society is not ready to talk about these disparities.

  164. Nick Says:

    I just want to share a small local news tidbit in case it becomes of interest later.

    It’s around 11:30 AM as I write this in Minneapolis. Word is going around on Facebook that there will be a march at 3:00 PM starting from US Bank Stadium led by Colin Kaepernick. This information got to me from my GF, who got it from a friend, who got it from an acquaintance.

  165. Gerard Says:

    anon #163

    I’m not even referring so much to that level of genius.

    I’d just like to understand why some people can score 700+ on the old-style SAT at 18 or younger without really trying or get easy A’s in challenging college calculus courses while others struggle greatly with those things, despite being highly intelligent in other ways.

    There’s a large enough sample of such people and it would be interesting to know what specific cognitive functions their enhanced mathematical ability correlates with as well as how it arises. For example is it mostly genetic or partly tied to early life interest in say playing with numbers versus reading books or even related to parental involvement.

  166. Scott Says:

    Gerard #165: Yeah, I’ve wondered about something similar. If we consider Scott Alexander, for example, I’d say that his general level of intelligence and insight is among the highest of anyone I know—and I know many Turing Award winners, Fields medalists, and Nobel laureates in physics. And yet the other Scott swears up and down that he struggled with basic math in school. How is such a thing possible?

    Part of the answer, as you said, might be that mathematical intuition is partly pre-verbal (e.g. visual and spatial)—but OK, what about the verbal parts? But I suspect that part of it is simply that, to become great at math, you have to really like staring at a piece of paper, looking for abstract patterns, distinguishing which of the patterns are merely “obvious” and which ones you can actually prove—enough to do it for years and years. And even among people with extremely high general intelligence, not everyone is going to like that enough—especially if they didn’t get positive feedback about it from early in life. Such people, alas, might have to settle for non-mathematical careers, like world-famous writer or philosopher or tech CEO. 🙂

  167. anon Says:

    Gerard #165:
    You want an uber-reductionist explanation for something that’s way beyond our understanding of genetics and personalities.

    Food for thought:

    12.5% of the fathers and 21.2% of the grandfathers (both paternal and maternal) of children with autism were engineers.
    Findings of this nature have led to the coinage of the term “geek syndrome”

    Wondering about the probability in the order direction 🙂 Also The Parable of the talents [StarSlateCodex]:

    Meanwhile, there were some students who did better than I did in Math with seemingly zero effort. I didn’t begrudge those students. But if they’d started trying to say they had exactly the same level of innate ability as I did, and the only difference was they were trying while I was slacking off, then I sure as hell would have begrudged them. Especially if I knew they were lazing around on the beach while I was poring over a textbook.


    But I still feel like there’s something going on here where the solution to me being bad at math and piano isn’t just “sweat blood and push through your brain’s aversion to these subjects until you make it stick”. When I read biographies of Ramanujan and other famous mathematicians, there’s no sense that they ever had to do that with math.

  168. PublicSchoolGrad Says:

    This entire post assumes that only entry into the top elite institutions counts. So what if you can’t into Harvard? You’ll be able to get into LSU, one of the Cal States or some state college somewhere. Your life is not over.

  169. Scott Says:

    PublicSchoolGrad #168: As someone who did his undergrad at Cornell (because that was where he was accepted), who later became a professor at MIT (which had actually rejected him twice, for undergrad and for grad school), and who now’s a professor at UT Austin (where he’s mentored several undergrads whose successes would be envied by most MIT students) … well, I couldn’t agree more! In fact, the more meritocratic the system is, the harder it is for us at UT Austin to compete against MIT and Stanford. We’re super grateful for their mistakes!

    But, like, does the undoubted fact that so many people succeed despite unfairness mean that we should be happy about unfairness, that we shouldn’t complain about it or try to fix it? Was it fine that all those Jewish kids were excluded from the math department at Moscow State University because of their surnames, since they could (and in the event, did) get a perfectly serviceable education at the Siberian Institute of Mining or wherever?

  170. PublicSchoolGrad Says:

    > In fact, the more meritocratic the system is, the harder it is for us at UT Austin to compete against MIT >and Stanford. We’re super grateful for their mistakes!
    They “system” consists of more than MIT, Stanford and the like. If the system is more “meritocratic” as you say, why would that make it harder to compete against MIT and Stanford? That assumes that UT Austin is getting some of the better students who would otherwise have gone to those places but only go to UT because they could not get into MIT/Stanford etc due to the latter’s non-meritocratic admission process. It also assumes that UT’s admission process is more meritocratic than MIT/Stanford/Berkeley etc.

    Incidentally, maybe it would better for the country as a whole if the the top tier places become even less meritocratic. That way, places like UT and other public schools would get the better students that could not get into the top tier places and this would improve the quality of the second tier places! I’m joking, of course.

    > But, like, does the undoubted fact that so many people succeed despite unfairness mean that we should > be happy about unfairness, that we shouldn’t complain about it or try to fix it? Was it fine that all those >Jewish kids were excluded from the math department at Moscow State University because of their >surnames, since they could (and in the event, did) get a perfectly serviceable education at the Siberian >Institute of Mining or wherever?

    Of course not and that’s not my point. All I’m saying is that if you’re a bright high schooler with motivation and poor grades/”leadership”/volunteer activities, your life is not over if you can’t get into UC Berkeley or whatever. There are probably hundreds of other fairly good universities around the country that you can get in and get an excellent education, even if they don’t have the name recognition or social capital of the Ivies or the “better” parts of the UC system. You may not get into Berkley but you may get into Davis or Irvine or one of the Cal States.

    I am sure you also realize that the current system has its own shortcomings. Your UT Austin, for example, is supposed to take the top 10% or 7% high school grads from Texas as a “fairness” measure but I’m pretty sure that the top 10% students at Sharpstown High School (ranked #1602 with over 95% free lunch recipients: do not get the same education as the top 10% at Westlake High School ( The inequity is built into the system from the ground up.

  171. Anonymous Ocelot Says:

    PublicSchoolGrad #170

    Re: the first half of your comment: I think you should reread #169 more closely… Scott is not wrong, he made the exact same joke you go on to make! I think you interpreted it the opposite way.

  172. fred Says:

    Scott #166

    “to become great at math, you have to really like staring at a piece of paper”

    Also a prerequisite to be a great writer or a great painter!

  173. fred Says:

    Eventuall candidates will be ranked based on their behavior inside immersive VR scenarios.
    Those will mix real-time problem solving and complex social interactions.

  174. Scott Says:

    fred #173: For mathematicians, a VR simulation of staring at a piece of paper doing math.

  175. John Figueroa Says:

    I feel the same way you do, Scott, so I don’t know if I can cheer you up. The United States clearly isn’t what I thought it was prior to November 2016. But I can share the brightest two offsetting factors that I’ve been clinging on to.

    First is the obvious one: the world and the United States have been through much worse, but things still got better afterward. Of course, the destructive power of humanity continues to expand, which tempers this light somewhat.

    But secondly: yeah, this might be the collapse of the American Leviathan, especially if he wins again. And yeah, Trump-minded people have been gaining ground in alleged liberal democracies everywhere. But there are pockets of hope. Germany and France in particular have sane, clear-minded, globalist, technocratic, liberal, democratic leaders. As much as I’m afraid of hope these days, the success of Macron’s En Marche! movement proves that globalist technocratic liberal democracy isn’t an intrinsic oxymoron. Of course, there’s been the populist backlash, and his favorability numbers aren’t great right now. But at least Merkel remains broadly popular. While the EU is floundering in its efforts to push back against authoritarianism within its borders, she, Macron, and others are working toward a saner, more united Europe.

    (I realize, of course, that the above is horrifying in a way. I don’t have a ton of hope for France’s next election. So maybe this all relies on the people of Germany successfully rejecting a budding right-wing populist backlash over the next year and a half.)

  176. John Figueroa Says:

    [PS: I messed up the link in my comment; this is what I meant to link to: Are the Germans Edging Closer to True Fiscal Union? In a striking reversal, Merkel joins with France in recommending a euro fund that could be a timid first step toward greater integration.]

  177. Scott Says:

    John Figueroa #175: Given everything that I’d learned about the world as a child, I agree with you about the historic irony that a good fraction of one’s hope for the future of liberty, reason, and Enlightenment on earth now rests with Germany.

  178. John Figueroa Says:

    Yeah… And it’s unbearably sad, in a way I can’t put into words properly, how little of my hope lies in America.

    One more thing that might indirectly help make you feel less depressed: acting depressed isn’t very effective. Spreading despair is a tactic, because despair is exactly what the enemy wants us to feel. I’m not advocating self-deception, just saying that doing the things a non-depressed person would do is probably better at reducing the things making us depressed than what we’re doing right now. (And from what I’ve read, as well as anecdotally, intentionally trying to do the things a non-depressed person does helps you feel less depressed).

  179. Timber Says:

    The GOP is the party which has been telling us since GOLDWATER!!!! that government is the “problem” – the enemy of happiness and efficiency. So it would be unreasonable to expect them (Trump, McConnell, any of them) to show any interest in good governance now, when doing so would only disprove their long-standing hypothesis.

  180. STEM Caveman Says:


    I was in a very similar position to you in high school and fully agree that tests are the great and wonderful gates of opportunity rather than the evil gatekeepers. ETS are the good guys as long as the tests are not too expensive. Grades, letters, interviews, are terribly inefficient and often disturbingly political, and any criticism of tests applies even more to the other criteria.

    With that said, you do a disservice by repeating cliches about ethnic or class competition, either the historically documented or present hypothetical kind, as the rationale for the current system. Someone building an admission system from zero, optimized to select the most driven, creative, high IQ geniuses-of-the-future would construct something not unlike the Harvard admission system, which (for STEM applicants competing in the academic track) is not much different from the Caltech or MIT system. Annoyingly and inefficiently collect all manner of second, third and fourth-order information about the candidates and try to sift them all.

    A couple of other points ignored in your post.

    One, that admission of students 2-3 years younger than the norm is (as it should be) a much harder case to justify. You would have been an automatic, rubber-stamp admit anywhere at age 17 assuming you continued on the “budding STEM god” path, so no point pretending any rejections at 14-15 are informative about admission in the normal route. I don’t know of many cases of prodigies who did well by going to the super-elite liberal arts universities years ahead of schedule. The successful ones went to tech schools where there was less need to adjust socially, or to lower ranked non-tech schools with minimal academic pressure to add to the growing pains.

    The other thing is that as education keeps getting stretched out for more and more years on average, and all the super-bright academically inclined kids end up in PhD programs, it is rational for universities to reduce the weight on IQ in undergraduate admissions. Stanford can recruit the geniuses for PhDs, and focus on other goals when selecting undergrads. I think this is what they and other highest ranking schools have in fact been doing and explains some of the patterns you may have seen or heard about.

  181. Max Says:

    People will always find a way. Even if your analysis is 100% correct there is always a way forward. United States had been a leader of a free world for much of XX but even if (big if) it slides into irrelevancy now it’s not such a big deal from the world point of view. There is always someone to pick up the torch.

  182. Erik Caine Says:

    I was scoring college graduate equivalents on standardized tests when I was 12 and from a very disadvantaged family in a very disadvantaged region of the country. I credit standardized tests as saving me from a likely horrible existence. I worked long hours in high school and I detested high school in any event. I was able to receive scholarships and grants and a private engineering school education primarily because of standardized tests. My work professionally has added significant value to many other lives and to society as a whole so I hate to see their demise. On this we agree-I think it is a mistake that wears further on meritocratic society and unevens the playing field that optimizes societal benefit..

    On the other hand when I read through the posts here I often see what seems to me an unwarranted conformity of thought that must (again it seems to me) be based on some shared ideological biases. As an example I was shocked to my core to see how little research has been conducted on transmission of Coronaviruses in aerosols (as well as other viruses) when I began looking at papers at the start of the pandemic. This is the third novel variant in the last two decades and with combined budgets of nearly $40 billion per annum and with recognized very high viral risk profile it was decided to terminate all ongoing research? The agencies tasked with preventing and/or mitigating pandemics of this type (and with handsome budgets) know nothing about potential asymptomatic transmission nor effectivity of masks with respect to aerosols and completely reverse their recommendations from time to time. When I saw the early press conferences I realized that those responsible for huge sums of research dollars knew zero about the small amount of Coronavirus research that was continued in Singapore and Hong Kong. The later modeling then completely ignored the high quality data available from the Diamond Princess and had an obvious huge bias when compared to that high quality data. I could go on about epidemiological indications that supported laboratory measurements showing temperature and humidity being pertinent to viability of Coronaviruses (structurally this variant is very similar to SARS. My conclusion is that to blame the course this pandemic took on Trump is mainly due to simplistic political ideology and misses the deep problems that need to be addressed.

    It seems almost childish to demagogue Trump to absurd lengths and I guess it is some kind of statement of solidarity. The obvious question for me is how two deeply flawed men could be the outcome of the political process in the US. Whoever wins the election will gather so many votes of people just voting against his opponent. A lot of Trumps support is a reaction to the left’s frequent arrogant claims of primacy of position based on scientific facts when no such facts exist. They are far too often simply ideologically based beliefs. As someone else pointed out he is a symptom rather than the cause of the problems everyone recognizes in the US.

    Your video recognized that criticism of groups in the US is part of healthy discourse and so I will add the following. It is undeniably true that the European ancestry Jewish people produce an incredible number at the highest reaches of human intelligence. Jewish physicists working in the US have contributed a staggering amount to the technological advances of global civilization in the 20th Century. However it seems to me that for social issues they have tended to support on average the usual left of center simple political solutions to problems that cannot be effectively addressed in the prescribed manner. I don’t know if it is a feature of traditional education, or the Torah, or some cultural survival mechanism from the terrible oppression that has been visited on them through recorded history. As best I can gather in Academia now you have to constantly demonstrate your ideological purity now or be cast out so maybe it is a cultural survival mechanism. Scientists ideally do science and should not be the reliable go to support resource for ideologists.

    When I look through history being on the side opposite mob enforced standards is the right side so I will hold my nose and vote for Trump not happy with the man but happy with what he opposes and who he is hated by. Sad that we are now in this position but that is where we are.

  183. Sam Says:

    With regards to the “university admissions” question, it seems to me that it might be useful to examine some extreme points on the distribution of admissions policies: we could start, for example, with the University of Warwick’s mathematics department which, as a matter of policy, makes its decisions purely on academic merit: they do not read any personal statements you send them, but instead simply set some level of academic achievement (in each education system that they’ve got applications from) and blanket accept everybody above that line, and nobody below it. Notably, for domestic applicants, this is entirely based on standardised tests (though not as they’re commonly known in the US: the STEP paper component in particular, is worth a look at – if nothing else, because there are genuinely interesting questions in there that are fun to think about; but even the A-levels are a large step away from the multiple-choice-dominated tests that seem to dominate discussions of standardised testing in the US).
    This contrasts rather starkly with other UK universities, which use something closer to the current US model (a mixture of a personal statement and academic achievement, though with the latter being given perhaps more weight, and being much less strongly based on subjective evaluations). The question, then, is what the differences between Warwick’s maths department and those other universities are, in terms of student body composition.
    From what I can tell (and admittedly, I don’t have good public data to post for Warwick, so I’m instead going off my own records and recollections of students from when I worked there), the answer is “not a massive amount”: in particular, the percentage of, for lack of a better term, “awkward nerds” in Warwick’s maths student body seems to be pretty much identical to that of Imperial (the closest academic parallel, but with a more “normal” admissions policy), and differences with other universities seem to be broadly in line with “more nerds at better universities”.
    More specifically, the percentage of students with special needs at Warwick was essentially identical to that of other universities that I’ve worked at.
    There also doesn’t seem to be much difference in diversity (the total spread of offer rates between ethnic groups in the UK has narrowed to almost nothing over the last decade:, although clearly were I writing this a decade ago, the answer would be have been very different. From my horribly unscientific observations, Warwick’s spread looks fairly similar to what we get on average, with the black population slightly underrepresented, probably due to education disparities in schools.
    Warwick’s maths department is reasonably heavily male-skewed, but it’s not an outlier in either direction from the general spread (we have universities like York with majority-female undergraduate maths cohort one one end, all the way through to universities with significantly larger male majorities in their maths departments).
    On poverty, however, the question is somewhat different: if you look under “Free School Meal status” (“Free School Meals” is a program that supports children of parents receiving certain means-tested government benefits – it roughly maps to an annual income of less than ~£16k) in that data, you’ll notice an absurdly wide split between the two groups. Sadly, Warwick also has a fairly wide split: the UK’s problem with income-based disparities in pre-university education aren’t as bad as those of the US, but they’re still pretty damned bad. When I worked there, we were making a big deal out of how we were doing better than the average (though it seemed to me that they were rather over-hyping “we’re still not doing a great job”). In particular, flat achievement-based entry criteria didn’t do *worse* than a hybrid method, so it seems that the attempts at correcting for disparities in opportunity made by other universities is at least balanced, if not outweighed, by the biases introduced by dealing with personal statements at all.

    All in all, it seems like there’s not a whole lot of difference across the spectrum of admissions criteria from “normal UK” to “pure testing”. The easy-to-find-figures on among the above (specifically race and poverty) seem to be worse in the US from what I can see (though admittedly, that might be distorted by the vastly different funding structures for undergraduate degrees), so (unless anybody from the US wants to share experiences as a point of comparison?) I’d guess that there probably is a difference between the US “mostly holistic” approach and the UK’s mixed approach, and not a favourable one for the US.

    Anybody have knowledge of any other extreme points that might be useful for bracketting the spectrum? Something like a US university that already doesn’t use any standardised tests? Or along some other axis?

  184. A. Karhukainen Says:

    Tangential to this discussion, from T. Greer’s blog: