Cracking the Top Fifty!
I’ve now been blogging for nearly twenty years—through five presidential administrations, my own moves from Waterloo to MIT to UT Austin, my work on algebrization and BosonSampling and BQP vs. PH and quantum money and shadow tomography, the publication of Quantum Computing Since Democritus, my courtship and marriage and the birth of my two kids, a global pandemic, the rise of super-powerful AI and the terrifying downfall of the liberal world order.
Yet all that time, through more than a thousand blog posts on quantum computing, complexity theory, philosophy, the state of the world, and everything else, I chased a form of recognition for my blogging that remained elusive.
Until now.
This week I received the following email:
I emailed regarding your blog Shtetl-Optimized Blog which was selected by FeedSpot as one of the Top 50 Quantum Computing Blogs on the web.
https://bloggers.feedspot.com/quantum_computing_blogs
We recommend adding your website link and other social media handles to get more visibility in our list, get better ranking and get discovered by brands for collaboration.
We’ve also created a badge for you to highlight this recognition. You can proudly display it on your website or share it with your followers on social media.
We’d be thankful if you can help us spread the word by briefly mentioning Top 50 Quantum Computing Blogs in any of your upcoming posts.
Please let me know if you can do the needful.
You read that correctly: Shtetl-Optimized is now officially one of the top 50 quantum computing blogs on the web. You can click the link to find the other 49.
Maybe it’s not unrelated to this new notoriety that, over the past few months, I’ve gotten a massively higher-than-usual volume of emailed solutions to the P vs. NP problem, as well as the other Clay Millennium Problems (sometimes all seven problems at once), as well as quantum gravity and life, the universe, and everything. I now get at least six or seven confident such emails per day.
While I don’t spend much time on this flood of scientific breakthroughs (how could I?), I’d like to note one detail that’s new. Many of the emails now include transcripts where ChatGPT fills in the details of the emailer’s theories for them—unironically, as though that ought to clinch the case. Who said generative AI wasn’t poised to change the world? Indeed, I’ll probably need to start relying on LLMs myself to keep up with the flood of fan mail, hate mail, crank mail, and advice-seeking mail.
Anyway, thanks for reading everyone! I look forward to another twenty years of Shtetl-Optimized, if my own health and the health of the world cooperate.
Follow
Comment #1 May 8th, 2025 at 5:01 pm
In a world of 8,000,000,000 people, being Top 50 at anything is quite an accomplishment!
Comment #2 May 8th, 2025 at 5:11 pm
LLM’s driving professional crackpots further into crackpottery is a big problem. I’m afraid it is also driving non-crackpots into crackpot adjacent crackpottery or amateur crackpottery. The recent ChatGPT sycophancy issues don’t help. https://www.vox.com/future-perfect/411318/openai-chatgpt-4o-artificial-intelligence-sam-altman-chatbot-personality
Turns out people love being told by an LLM that there views on the world are right, correct, wonderful, original, and amazing!
Comment #3 May 8th, 2025 at 5:49 pm
Adam Treat #2
that said, I also wonder what the world will be like once/if AGIs are a thing, capable of independently discover new facts 24/7 and turn truth into a commodity as abundant as water.
Like you open the AI tap for one second and get overwhelmed with a 100,000 pages of dense mathematical/scientific/engineering, a single one of which would take us 4 months to grasp/understand as a mere human, even with the help of the most advanced AI assistant… we’ll just become very quickly irrelevant. Sure, at first, we’ll give them very specific human oriented goals, but once those goals are met, we once again will run into the limits of our own understanding and limited imaginations as humans.
It will be as if, as humans, we manage to communicate with dogs, and then we let them tell us how to run a city, by setting goals and strategies (make more treats! make more doggie parks!). Quickly the dogs would be content, lose interest, and be confused as what to do next, and would let us run the show while they do their dog things.
Every single species on the planet sees themselves as “humans” (every organism is at the center of the universe, with its own perceptions, by definition), with species above them and species below them… it’s just that so far we’ve been the only species with no other species above them, but this is about to change. We’re in for a rude shock.
Comment #4 May 8th, 2025 at 7:19 pm
The use of AI is now highly visible on technical discussion forums too. In the realm of physics, I regularly see people posting theories of everything or ideas for new physics, in which AI has supplied some or all of the details. It also happens on Less Wrong – new accounts posting essays on metaphysics, alignment, etc. On Q&A sites, it’s certainly common for AI to be used to generate answers, and the telltales of this are getting harder to detect. Then one also has non-English speakers who are using AI for anything from fixing their prose, to writing up to 100% of the whole thing.
The thing is that AI really can do research (ChatGPT, Gemini, Claude all now have Deep Research options I think), and even generate new ideas (just by combination), as well as e.g. translate a paper and analyze it (I had occasion to do that via o3 recently). One could spend a lifetime investigating all the different forms of AI-assisted writing and cognition that we are seeing.
When someone just asks an AI for a new physics idea, and gets back e.g. some hypothesis about dark energy, that they send to a human scientist or post on a forum… this is a new kind of human-computer interaction, that could be studied and characterized. There’s the psychology of the human who is seeking answers and treating the AI as an oracle; there’s the process whereby the request interacts with the labyrinthine pathways of the AI’s cognitive landscape, built up from some training corpus and partly explorable via interpretability techniques; the human response to the output which, if it passes muster, ends up being forwarded to a human expert or placed before the readers of a forum…
This cycle of events clearly has a lot in common with what scientific outsiders, or just independent thinkers (whether they are students or retired engineers), do when they post their own theories, but in this case, so much of the theory formation has been outsourced to the AI (and of course, how much gets outsourced varies – the human can have the idea and discuss it with the AI, or just ask it to write a short essay which gets immediately posted, or you can even just ask the AI for a new idea). This is enough of a phenomenon that even vixra, the alternative to arxiv, now has a special subcategory specifically for AI-generated papers.
AI-human cognitive symbiosis is already a reality (and in far more voluminous form, e.g. all those people using AI to write their CVs and filter their work email), and it surely already takes more forms than any one person grasps (though presumably there are people at places like OpenAI, who try to keep up with what their users are doing with AI, who would have a surprising breadth of insight into the multifarious ways in which AI is being used). And coming up are what I call “research swarms” in which the AIs are talking to each other – this is a large part of the “AI 2027” scenario – and here we should expect that the modes and rationales of AI-to-AI communication will even more exceed human comprehension, even if there are agents dedicated to making a daily summary for their human keepers, of the latest trends in the cognitive hive.
Comment #5 May 9th, 2025 at 4:58 am
Congrats and thank you. 🙂
It’s immeasurable how much your blog has influenced and helped me in my academic and professional life so far already, and I am very much looking forward to the twenty years to come. May there always be a Shtetl post in my inbox when I am pondering over another quantum news.
Comment #6 May 9th, 2025 at 5:50 am
AI bots and AI generated ‘content’ is already invading social media, taking away business from influencers etc. and some are already speculating that AI will destroy social media and the web as we know it.
I think this may be the best thing to happen in a long time.
Comment #7 May 9th, 2025 at 6:05 am
Hi Scott,
I can’t claim to be a long-time fan. I only discovered your blog shortly after the horror of October 7. While I was grappling with my sanity, when many of my fellow students (who called themselves anti-colonialists, anti-imperialists, etc.) celebrated the largest mass slaughter of Jews since the Holocaust, your blog was (and I’m not exaggerating or being dramatic here) a light in the darkness. You gave me some hope after that dark day. And your moral and intellectual courage inspired me not to abandon my love for Israel and the enlightenment project more generally.
Then I discovered some of your writings about mathematics, physics, logic, and theoretical computer science, and they were wonderful. I can’t claim to know much about quantum complexity theory—never studied it, and it’s not on my agenda—but you write plenty about broader stuff (most recently foundations of math and how it relates to computability) that are super interesting to me as a math/physics guy. In particular I enjoyed one of your “early writings” about the Busy Beaver function, “find the largest number” 🙂
You and your writings are truly unique. You are probably the most open-minded guy on the internet, most willing to debate any position on any topic, explore intellectual assumptions that nobody else would dare question. This place truly feels like the agora of Athens at the time of Socrates or Plato, and that’s a rare and wonderful thing.
And you’re a kind person. You helped encourage me to get outside my comfort zone and join some social clubs at my uni. Few academics would spend their time to help a stranger. Sorry for this really long comment, but I know you’re coping with cruel bullies and trolls in your comment section, so I just wanted to tell you that you’ve got plenty of fans as well 🙂
Best wishes,
Julian
Comment #8 May 9th, 2025 at 6:24 am
“Anyway, thanks for reading everyone! I look forward to another twenty years of Shtetl-Optimized, if my own health and the health of the world cooperate.”
And we hope to be here to see it. Thanks so much for this half trip.
Comment #9 May 9th, 2025 at 7:37 am
I had kinda assumed yours was the top quantum computing blog on the web, or at least in the top few such blogs. Marking “top-50” seems very – (checks the list) – oh.
Comment #10 May 9th, 2025 at 9:22 am
Top 50? That’s nothing. This is one of the top dozen blogs on my all-time list. I began reading it almost 20 years ago. I won’t be able to make it through another 20, but I’m happy to have had as much of it as I have.
Comment #11 May 9th, 2025 at 10:03 am
Congratulations Scott, but don’t sell yourself short. You’ve been at the pinacle of the quantum blogosphere since 2007, when Ricoh used fashion models plagiarising your work to sell printers 😉
https://scottaaronson.blog/?p=277
Comment #12 May 9th, 2025 at 11:13 am
You were the top blog for a long time, when QC was super niche, until the hype train left the station and everyone and their grandma became experts overnight.
Comment #13 May 9th, 2025 at 1:05 pm
fred #12:
everyone and their grandma became experts overnight
Did they, though? 😀
Comment #14 May 9th, 2025 at 2:58 pm
Congratulations, Scott. I’ve been following your blog for years and consider it better than one of the Top Fifty as it is, to my knowledge, unique.
You discuss your subject as someone who understands it thoroughly, and more than that you discuss it with humour and compassion and a transparent openness. In a world of fakeness, hypocrisy, cynicism and violence you shine with the simple light of human decency.
Lang may yer lum reek,
David
Comment #15 May 9th, 2025 at 5:15 pm
Thanks so much, everyone!!
Comment #16 May 9th, 2025 at 5:58 pm
1) I thought you were the Number 1 quantum blog (or since its hard to compare, perhaps the top 5).
2) I am surprised you are only getting the recognition now, though perhaps the list itself is new.
3) Lance is also getting more emails with proofs that resolve P vs NP. He didn’t mention ChatGPT or AI as a cause, but that is quite plausible.
4) Someone claimed to crack all the Clay Math Problems– well, if you prove P=NP perhaps you can use that to generated proofs of the rest.
Comment #17 May 9th, 2025 at 8:48 pm
William Gasarch #16:
I thought you were the Number 1 quantum blog (or since its hard to compare, perhaps the top 5).
I was hoping the commenters would say that, not me! 😀 But yes, of course the first thing I wondered was what the other 49 were…
Comment #18 May 10th, 2025 at 3:52 am
This was surely the first and only blog about quantum computing 20 years ago. Now it seems there are even 50 blogs on QC out there. Scott, if 20 years from now you make it to the top 1000, what a legacy it will be!
Comment #19 May 10th, 2025 at 5:53 am
When I was a teenager and a physics nerd I was fascinated with crackpots and their theories of everything. There was a great book, Physics on the Fringe, about these people and why they’re so attracted to crank ideas. I recommend reading it if you want insight into these cranks.
One of the most hilarious videos I’ve ever seen was “tales from the box,” where this physics professor from Cal Poly showed off some of the crank mail the department had received over the years. If you want a laugh you should watch it!
I have this idea: How about dedicating a blog post to these cranks, showing us the funniest and weirdest theories that landed in your inbox? It might be a bit too mean-spirited for this blog, but I think it could be really funny and help you with your crank frustration.
Cheers,
Julian
Comment #20 May 10th, 2025 at 7:15 am
I think it goes without saying that you’re the best QC blog on the internet. Or at least I can say, since I’m not a QC researcher, that you’re the best blog bridging the gap between QC and a popular audience. And indeed, probably the best one bridging the gap between CS theory and a popular audience!
Regarding the proofs, I saw some math prof on Twitter mention that he’s getting a ton of crank proofs. Don’t remember who, and I don’t think he explicitly said that it’s because of LLMs but that was everyone’s theory.
Comment #21 May 10th, 2025 at 12:35 pm
Way to go! But I don’t see the badge proudly displayed here — I take it you just haven’t gotten around to adding it yet?
If I’m reading this right, you’re ranked #18 by “domain authority” with a score of 62. #17 WRAL TechWire has a score of 69. #1 Microsoft Quantum Blog has a score of 99.
Hang in there! Keep hustling and grinding, and some day you just might crack the top ten.
Comment #22 May 10th, 2025 at 5:03 pm
Scott #17:
You’ll never believe #7!
Comment #23 May 11th, 2025 at 4:31 am
You are now and have clearly been the #1 top Quantum Computing blog for many years.
Your anonymous regular (too regular) poster “fred” is particularly annoying with the pretty dumb posts #1 and #12.
I don’t post much anymore since I have more important family concerns and I also am quite unsettled about how crazy the world has become.
Comment #24 May 11th, 2025 at 5:26 am
I recently learned that the kerr solution has closed timelike curves. It happens in the so called negative mass universe (where r<0), you get to a point where g_phiphi<0 and therefore phi becomes timelike coordinate, meaning there are closed timelike curves as phi is cyclic coordinate. The only downside is that you'll never be able to return to the original universe and report on whatever you found. It is also not known (it's still actively researched) whether you can make it through the Cauchy horizon.
Do you think the universe will allow you the computational power of CTCs if you reach inside rotating black holes? Or will something stop you before you reach the negative mass universe?
Comment #25 May 11th, 2025 at 8:39 am
Physics student #24: That’s an excellent question to which I don’t know the answer!
In the Kerr solution, is the CTC so close to the singularity that we’d expect quantum effects to change everything anyway, or can it have macroscopic size? In classical GR, if you modify the Kerr solution by sending some matter into the CTC, what if anything happens to prevent causality paradoxes? Happy to hear comments from anyone who knows GR (and is outside a Kerr black hole so we can receive a message from them)… 🙂
Comment #26 May 11th, 2025 at 10:09 am
Scott #25: “Happy to hear comments from anyone who knows GR”
The energy for a particle with a finite rest mass required to run through such a closed timelike curve is many times its rest mass, see for example my PSE question Energy balance of closed timelike curves in Gödel’s universe
Do I know GR? I certainly know more GR today than I knew back then in 2012. However, my knowledge for that specific question has not really changed. I can not even tell you whether “many times its rest mass” is true for any CTC, or just for CTCs in Gödel’s universe.
Comment #27 May 11th, 2025 at 12:37 pm
Scott #25:
There are CTCs in the Schwarzchild (and Kerr) solutions, but in any case, they require traversing the singularity.
There are no known solutions to the Einstein field equations (i.e., metric tensors on R^4 satisfying these equations) with positive mass-energy density with CTCs.
In fact, it’s a conjecture in GR theory that there are no CTCs appearing in any smooth solutions of the EFE with positive mass-energy density. This is an unproven conjecture, and any young nerd here can earn eternal glory by finding a counterexample! There are CTCs in negative mass-energy spacetimes (wormholes, of course) but not in positive case (that don’t involve traversing a singularity).
Comment #28 May 11th, 2025 at 2:43 pm
There are no closed timelike curves in the Schwarzschild solution, and the idea of a curve “traversing the singularity” makes no sense because singularities aren’t points in spacetime.
Comment #29 May 12th, 2025 at 4:40 am
Thank you, Scott Aaronson, for your article and for sharing this exciting milestone in your blogging journey, as well as your continued contributions to the world of quantum computing and beyond.
Comment #30 May 12th, 2025 at 6:12 am
Hi John,
My language was sloppy, I’m sorry.
I was imagining a situation along these lines:
1. A four-manifold M
2. A larger manifold M’ containing M as an open subspace
3. A pseudo-Riemannian metric on the manifold M
The example that was in my head was the Schwarzchild solution in Kruskal coordinates, in which case M is the manifold R \times R \times S^2 \ {r \leq 0} (r is a function of coordinates R, T, I’m using the notation from Carroll’s book). The larger manifold M is the space R \times R \times S^2 \ {r M’ such that for every t \in R, if f(t) lies in M, then it is timelike in the metric on M.
In our situation, the map f: [-1, 1] -> M’ which sends -1 and 1 to the identified point and for t \in (-1,1) to the point (T, 0, 0, 0) is timelike by this definition, and closed.
My intuition here was that I was trying to replicate something I had heard (or I believed I heard?) long ago about travelling from the infinite past to the infinite future from a white hole to a black hole. There was something about a CTC in Kruskal spacetime, at least I thought I heard that somewhere, that’s why this situation came to mind.
The problem though is that this definition of a CTC on an extended spacetime with singularities doesn’t seem useful, isn’t very interesting.
In particular, I was disappointed to discover that, by this definition, there are CTCs on Minkowski spacetime.
Namely identify Minkowski spacetime as a metric on R^3 \times (0,1) = M, and pick M’ = R^3 \times [0,1] with all the points of the form (x,y,z,0) and (x,y,z,1) identified. Then we get a CTC in M’ by taking the map f:[0,1] -> M’ sending 0 and 1 to the identified point and sending t \in (0,1) to (0,0,0,t).
Maybe we can refine this concept a little more to avoid this. Namely, the extra points in M’ should be singularities of the metric on M. To be precise, let M be a four-manifold with a pseudo-Riemannian metric, and say M’ is a singularity-extension of M if
1. M is an open submanifold of M’
2. M’ \ M consists of a finite number of points (or a discrete set of points?)
3. Each point p \in M’ \ M is a singularity of the metric on M. Namely, for each such p, there exists a curvature scalar invariant K of the metric such that for any \sigma > 0, there exists an open subset U of M’ containing p such that within U \cap M, K > \sigma.
Then for a solution of the EFE with singularities, we can ask about CTCs on the singularity-extended spacetime. Perhaps we can call these “jumping paths,” because we can identify any of these singularities in M’.
Is this a useful concept at all? Is it interesting or does it end up with something trivial?
This is what I had in mind talking about paths “traversing singularities” in spacetime.
Comment #31 May 12th, 2025 at 10:08 am
Anyone out there have enough energy to check out the other 49? If so, perhaps you could provide us with an estimate how many:
1. Have some pretty good stuff,
2. Appear to have some idea what they are talking about,
3. Rather clueless,
4. Utterly clueless,
4. AI generated crap.
I am going to guess that 3, 4, and 5 will be the big winners.
I waste a lot of time reading a paragraph or so of all kinds of news items. A recurring theme in business and invertment publications is “should you put your money in a mixture of QC companies, or should you go for a ‘pure play’ and put everything into one QC company, often DWave”. This doesn’t give one a lot of confidence in the advice out there.
Comment #32 May 12th, 2025 at 2:35 pm
Hi Scott,
I was reflecting yesterday, and I think I identified a crucial difference between you and the woke.
Imagine reversing political orientations, and imagine that Peter Woit had said precisely the same things in a situation where mobs of students were harassing black people on campus, saying they should go back to Africa, etc. He would’ve been cancelled by the wokes. People would give me shit for even reading his book about Quantum Mechanics, called me a racist for it.
You, on the other hand (I hope! Am I right?) won’t give me any trouble for still being a fan of his work, for still being a fan of his Quantum book. I think his book on quantum mechanics was an excellent introduction to representation theory and QM. Unlike the wokes in the analogous situation, you won’t call me an antisemitic traitor for appreciating his book, even for being a fan. Even when you’re righteously angry (and for reasons I absolutely agree with!) about antisemitism, you don’t have the same “cancellation impulse” as the woke.
Comment #33 May 12th, 2025 at 2:41 pm
> Indeed, I’ll probably need to start relying on LLMs myself to keep up with the flood of fan mail, hate mail, crank mail, and advice-seeking mail.
Just for fun, I decided to get a LLM to quickly code up a classifier. (I didn’t even test it, so consider it more like pseudocode.) https://pastebin.com/6S59b0At
I like the test cases it came up with:
email_categories = [
“Fan Mail”,
“Hate Mail”,
“Crank Mail”,
“Advice-Seeking Mail”,
NONE_OF_THE_ABOVE_LABEL # This label will trigger a None return
]
# Example email texts
sample_email_1 = (
“Subject: You’re amazing!\n\n”
“Dear Creator,\nI just wanted to reach out and say how much I absolutely ADORE your work. ”
“Your latest video on quantum banana bread was revolutionary! It’s inspired me to start my own ”
“bakery that only accepts cryptocurrency. You’re a true visionary!\n”
“Keep up the fantastic content!\nYour Biggest Fan, Starla”
)
sample_email_2 = (
“Subject: YOU ARE AWFUL\n\n”
“To the person responsible for that last article,\nI can’t believe the utter nonsense you published. ”
“It’s clear you have no idea what you’re talking about. My dog could write a better piece. ”
“I’ve unsubscribed and told all my friends to avoid your content. You should be ashamed.\n”
“Sincerely, A Very Disappointed (Former) Reader”
)
sample_email_3 = (
“Subject: URGENT: The Squirrels Know!!!\n\n”
“Esteemed Recipient,\nI have uncovered a conspiracy of global proportions. The squirrels in my backyard ”
“are not what they seem. They are collecting acorns not for winter, but to power their subterranean ”
“mind-control device that will make us all crave broccoli! We must replace all bird feeders with ”
“tiny tinfoil hats IMMEDIATELY. There is no time to lose! Do not trust the pigeons, they are informants.\n”
“A Concerned Citizen, Bartholomew P. Quibble”
)
sample_email_4 = (
“Subject: Seeking your wisdom – Career Path Question\n\n”
“Dear [Your Name/Organization],\nI’ve been following your work for a while and greatly admire your ”
“insights in the field of [relevant field]. I’m currently at a crossroads in my career, trying to decide ”
“between specializing in advanced teapot design or pursuing a PhD in interpretive dance. ”
“Any advice you could offer on how to navigate this decision would be immensely appreciated.\n”
“Thank you for your time and consideration,\nAspiring Professional”
)
sample_email_5 = (
“Subject: Weekly Project Sync – Update\n\n”
“Hi Team,\nJust a reminder that our weekly project synchronization meeting is scheduled for ”
“tomorrow, Tuesday, at 10:00 AM PST via Zoom. The agenda includes a review of last week’s ”
“milestones and planning for the upcoming sprint. Please ensure your task updates are in JIRA.\n”
“Best,\nProject Manager”
) # This one is likely “None of the above”
emails_to_test = {
“Email 1 (Fan Mail)”: sample_email_1,
“Email 2 (Hate Mail)”: sample_email_2,
“Email 3 (Crank Mail)”: sample_email_3,
“Email 4 (Advice-Seeking Mail)”: sample_email_4,
“Email 5 (Work/Other)”: sample_email_5,
}
Comment #34 May 12th, 2025 at 3:14 pm
Actually, Claude’s code is way nicer and easier to read: https://pastebin.com/eUz0Yr2p
Comment #35 May 12th, 2025 at 3:14 pm
Julian #32: Indeed, I will not cancel anyone for having any opinion they choose about Peter Woit’s writings on representation theory and QM. 🙂
Comment #36 May 12th, 2025 at 3:28 pm
I’m curious how you feel about the adoption of “woke” cancel culture techniques by the pro-Israel right.
In particular, groups like Betar doxxing students who are in organizations like CUAD (which is, to be fair, a truly odious organization), but also students who just wrote op-eds critical of Israel, and finding Israel critics on the street and following and videotaping them, etc.
I get where they’re coming from. Many of these people really are odious and I understand the rage after October 7.
At the same time, I’m a huge critic of wokeism and cancel culture and I think adopting their techniques and their intellectual culture to combat antisemitism is a huge mistake.
What do you think?
Comment #37 May 12th, 2025 at 4:07 pm
I also couldn’t resist having ChatGPT design the notifications.
https://chatgpt.com/backend-api/public_content/enc/eyJpZCI6Im1fNjgyMjYxYzYyY2ZjODE5MThkMTgzZmMxZTRkMjYyMGQ6ZmlsZV8wMDAwMDAwMDU0Y2M2MWZkYTQyNDc2ZDdmZDVlNjg3NiIsInRzIjoiNDg1MzAxIiwicCI6InB5aSIsInNpZyI6IjM5ZDRmMTE2MTNkYjJmMThhODYxZWY5MTgwYzgxYjFiNTQyNjU3MzVjNjQ2ODMyYjgzMDhjZmY0NTU0MjVjMDAiLCJ2IjoiMCIsImdpem1vX2lkIjpudWxsfQ==
Comment #38 May 12th, 2025 at 4:19 pm
Julian #36: I haven’t followed Betar’s activities in any detail but if those who’ve done the most to defend Zionism in academia, like Shai Davidai at Columbia, have denounced Betar for going too far then they probably have very good reasons!
Comment #39 May 12th, 2025 at 4:38 pm
I see.
This might be a too-general question, and I’m sorry if this is veering too far off topic for this post, but: How do you feel about “cancelling” academics, for example, for antisemitism? For instance, revoking speaking engagements all the way up to firing tenured faculty, for anything from criticizing Israel all the way to celebrating the murders of October 7? Or revoking students’ job offers for participation in Tentifada? Would it be better if harsh critics of Israel in academia kept their mouths shut, through fear if necessary?
Comment #40 May 12th, 2025 at 7:56 pm
Julian #39: In general, I think universities ought to offer extremely wide latitude to express unpopular opinions, even opinions that outrage or offend people, and that many consider racist or antisemitic or the like. At the same time, there need to be time, place, and manner restrictions: taking over a building or shouting down a speaker isn’t pure “free speech,” it’s restricting others’ freedom to speak or be heard or work or study (which, indeed, is typically the point).
There’s also a serious problem if, for example, professors act in a discriminatory manner toward Jewish or Israeli students, or if student clubs systematically exclude those students. Again, this goes beyond expression of controversial opinion, and creates a hostile environment that de facto excludes certain students because of who they are.
So, I think universities need to have rules about these matters, and then actually enforce their rules in an evenhanded and content-neutral way. There’s of course a lot more to say about these matters, but maybe that’s enough for a blog comment.
Comment #41 May 12th, 2025 at 9:49 pm
Scott #40
> So, I think universities need to have rules about these matters
Oh, boy, do they ever!
> and then actually enforce their rules in an evenhanded and content-neutral way.
Ah.
Comment #42 May 13th, 2025 at 6:34 am
6 or 7 such emails per day? That’s a lot! Actually I’m curious about their contents, especially the P Vs NP ones, curious how fast I would be able to spot the fallacy.
Comment #43 May 13th, 2025 at 6:02 pm
Some interesting questions / puzzles that came out of a discussion with a friend:
1) Does there exist a polynomial time algorithm that can answer two instances of SAT by making one query to an oracle for SAT? (Assume P!=NP, SAT is considered a decision problem)
2) Does there exist a polynomial time algorithm that can solve two instances of the halting problem by making one query to an oracle for halting problem?
3) Does there exist an algorithm that can solve two instances of the halting problem by making one query to an oracle for halting problem?
I don’t know the answer to the last one yet, the rest were quite cute / interesting to reason and prove. Wanted to share with folks in the comment section. Feel free to share solutions if not shared already. Is there an analogue of this question for other complexity / computability classes such as BQP which would be interesting?
Comment #44 May 15th, 2025 at 12:47 am
Former Student #43,
I think I solved them all. (Just now; therefore I could easily be wrong.) Here are my solution notes in rot13, to avoid early spoilers. Thanks very much for these, they were indeed quite puzzling and fun.
(To really avoid spoilers, don’t even skim it in rot13 form! That cryptosystem does have some disadvantages.)
> (1) Qbrf gurer rkvfg n cbylabzvny gvzr nytbevguz gung pna nafjre gjb vafgnaprf bs FNG ol znxvat bar dhrel gb na benpyr sbe FNG? (Nffhzr C!=AC, FNG vf pbafvqrerq n qrpvfvba ceboyrz)
V guvax V pna cebir lbh pna’g qb (1), ol svaqvat na nyt gb fbyir FNG va cbyl gvzr haqre gur nffhzcgvba bs (1).
Yrg zr fcryy bhg gung nyt. Fnl nyt N qbrf gur fgngrq guvat va (1) va gvzr c(a).
Gura ol vaqhpgvba va a, V’yy znxr n FNG nyt gung gnxrf gvzr B(a c(a)) (naq arrqf ab benpyr).
Vg jvyy fcraq B(c(a)) gvzr gb erqhpr n FNG ceboyrz K bs fvmr a gb n ybjre fvmr.
Gur xrl vf gb erjbex nyt N vagb n inevnag O juvpu tvirf lbh bar be gjb cbffvovyvgvrf sbe gur pbzob bs gur gjb fbyhgvbaf gb vgf gjb vachg ceboyrzf. (O jvyy fvzhyngr N naq cergraq rnpu nafjre sbe gur benpyr va ghea, ynoryyvat rnpu nafjre (n fbyhgvba sbe obgu vachg ceboyrzf) jvgu jung gur benpyr fnvq gb pnhfr N gb tvir gung nafjre. Juvpurire nafjre vf evtug, O’f nafjre ynoryyrq jvgu gung pnfr zhfg or evtug, fvapr O fvzhyngrq N pbeerpgyl va gung pnfr.)
Gura yrg K0 zrna “K unf n fbyhgvba jvgu vgf svefg inevnoyr frg gb 0” naq fvzvyneyl sbe K1.
Abgr FNG(K) = FNG(K0) Be FNG(K1).
Fb lbh nfx O nobhg gur gjb ceboyrzf FNG(K0), FNG(K1). Lbh pna vagrecerg O’f nafjre (2 cbffvovyvgvrf bhg bs gur 4) nf rknpgyl bar bs gur ybtvpny fgngrzragf
1- FNG(K0)
2- abg FNG(K0)
3- FNG(K1)
4- abg FNG(K1)
5- FNG(K0) == FNG(K1)
6- FNG(K0) != FNG(K1)
(hayrff obgu O’f cbffvovyvgvrf ner gur fnzr, va juvpu pnfr vg nyernql qrpvqrq FNG(K) sbe lbh, ol qrpvqvat obgu FNG(K0) naq FNG(K1), fb lbh ner qbar).
Vs O’f nafjre vf 1 be 3, gura lbh xabj FNG(K) vf gehr.
Vs O’f nafjre vf 2 (abg FNG(K0)), gura lbh xabj FNG(K) == FNG(K1), vr lbh erqhprq FNG(K) gb n fznyyre ceboyrz. Jr hfr bhe vaqhpgvir fgrc gb qrpvqr FNG(K1), gura jr’er qbar.
Fnzr jvgu 4.
Vs O’f nafjre vf 5, jr nyfb xabj FNG(K) == FNG(K0) == FNG(K1), fb cvpx rvgure bar naq erqhpr gb gung, qbar.
Vs O’f nafjre vf 6, FNG(K0) != FNG(K1), gura jr xabj FNG(K) vf gehr evtug abj, qbar.
DRQ.
> 2) Qbrf gurer rkvfg n cbylabzvny gvzr nytbevguz gung pna fbyir gjb vafgnaprf bs gur unygvat ceboyrz ol znxvat bar dhrel gb na benpyr sbe unygvat ceboyrz?
V xabj lbh (BC) qvfcebirq guvf, bgurejvfr lbh jbhyq abg pbafvqre (3) bcra.
V pna’g vzntvar ubj guvf pbhyq or qvfcebirq jvgubhg fbyivat (3), rkprcg ol hfvat bayl gur snpg gung gurer vf fbzr xabja naq pbzchgnoyr gvzr yvzvg sbe guvf nytbevguz.
Fb V qvq abg fbyir (2)… rkprcg gung zl fbyhgvba sbe (3) vzcyvrf n fbyhgvba sbe (2). Ohg gung zrnaf V unir ab vqrn ubj lbh fbyirq (2).
> 3) Qbrf gurer rkvfg na nytbevguz gung pna fbyir gjb vafgnaprf bs gur unygvat ceboyrz ol znxvat bar dhrel gb na benpyr sbe unygvat ceboyrz?
V pna cebir lbh pna’g qb guvf onfrq ba fbzr onfvp cebcregvrf bs gur Ohfl Ornire shapgvba OO(a). Vs nalbar vf abg snzvyvne jvgu gung, ernq bhe ubfg’f rkpryyrag fheirl.
Fhccbfr gurer jnf na nyt N jvgu guvf cebcregl. Nf va zl fbyhgvba gb (1), jr ghea vg vagb nyt O, juvpu fvzhyngrf N, ohg guvf gvzr O bhgchgf gur arj cebtenz gung N jnagf gb nfx gur benpyr nobhg, nf jryy nf jung nafjref N cynaf gb tvir sbe rvgure benpyr nafjre nobhg jurgure gung cebtenz unygf.
Abj pbafvqre nyt P juvpu vf tvira n ahzore a. Vg znxrf n ovanel gerr bs nyy 2^a cebtenzf rkcerffvoyr va a ovgf. Ng rnpu yriry bs gur gerr, vg hfrf O gb ghea gjb puvyq cebtenzf vagb bar cnerag cebtenz (creuncf zhpu ynetre, ohg gung qbrfa’g znggre). Riraghnyyl vg trgf n fvatyr cebtenz ng gur gbc.
Vs vg xarj jurgure gung cebtenz unygrq, vg jbhyq xabj jurgure rnpu bgure cebtenz va gur gerr unygrq (onfrq ba O’f bgure bhgchgf gurer), vapyhqvat nyy gur yrns cebtenzf. Vs vg xarj gung, vg pbhyq fnsryl fvzhyngr nyy naq bayl gur yrns cebtenz juvpu unyg, guhf ehaavat sbe ybatre guna nal bs gurz, ohg unygvat.
Va bgure jbeqf vg pbhyq pbzchgr OO(a) hfvat na rapbqvat bs a, bar ovg bs nqivpr (nobhg gur unygvat bs gur gbc cebtenz ba gur gerr), naq B(1) ovgf sbe nyy vgf bgure pbqr.
Fvapr nal ynetr rabhtu a pna or pbzcnpgyl rapbqrq (va B(ybt a) ovgf), guvf trgf n pbagenqvpgvba (sbe ynetr rabhtu a) ol svaqvat n cebtenz fubegre guna a juvpu pna pbzchgr OO(a). DRQ.
[raq]
Comment #45 May 15th, 2025 at 8:25 am
@bruce thankfully chatgpt is quite capable of translating rot13 cypher! I will not use rot13 below:
My solution to (1) is very similar to Bruce, but-proof of (2) and (3) are different. I solved (3) after I posted my previous comment. I think there’s a flaw in Bruce’s proof of (3): the algorithm A will not necessarily halt for both branches of computation from the oracle’s answer. If the oracle’s answer is incorrect, it may continue to do some computation that never halts. So it is not possible to simulate A to return both possible outputs without making an oracle query. This is where the runtime bound helps: it doesn’t have to be polynomial, but if we know some computable function which is a bound on the runtime of A, we can simulate both branches of A to provide possible answers. Without a runtime bound on A, it is still possible to reduce BB(a) to just one query to the halting oracle. But I didn’t understand how the advice bit can be encoded into the input to the BB() calculator, so maybe your solution can be fixed? Feel free to explain that part better.
So I will prove (3) directly in a completely different way below:
Suppose such an algorithm A exists. It takes as input m1, m2, constructs m3 = q(m1, m2), makes a single oracle query for the halting status of m3, depending on the output, does some arbitrary computation that calculates the halting status of m1 and m2. The remaining arbitrary computation only needs to halt assuming the oracle answered correctly. We will use this machine to construct another algorithm that can solve the halting problem without using any oracles.
Define f(m1, m2) = 0 if q(m1, m2) halts, Otherwise f(m1, m2) = 1
Define g(m1) = 1 if f(m1, m2) = 1 for all m2. Otherwise g(m1) = 0
Case 1: g(m) = 0 for all m. Construct an algorithm for the halting problem for machine m1: for n = 0 to infinity: for m2 = 0 to n (using some encoding of machines as integers): Calculate q(m1, m2) and simulate it for n steps. If it halts, simulate rest of algorithm A(m1, m2) to find the halting status of m1, m2 and output the status of m1. According to our assumption, q(m1, m2) will halt for some m2, and some n. So this algorithm will terminate and give correct answer.
Case 2: g(M) = 1 for some smallest machine M. Construct an algorithm for the halting problem for machine m1: simulate A(M, m1) assuming the oracle output will be non-halting (without making queries) and output the status of m1.
Whether case 1 is true or case 2 is true may be difficult to determine, but it only depends on A, and in both cases we have demonstrated an algorithm that can solve the halting problem correctly without making any oracle queries. Hence original assumption that A exists is false.
Comment #46 May 15th, 2025 at 8:47 am
Former Student #43,
About generalizing to other complexity classes, I think my solution to (1), for SAT (NP-complete), applies virtually unchanged to QBF (PSPACE-complete).
Comment #47 May 15th, 2025 at 10:38 am
Former Student #43,
I can simplify and generalize my solution to (3). The simplification is to remove its dependency on a certain other known result. Even to state the generalization would be a big hint about the original solution, so it’s all in rot13 for now.
Fhccbfr, sbe nal svkrq x > 0, gurer vf na nytbevguz N juvpu, tvira nal x cebtenzf C_v, bhgchgf x ovgf o_v, jvgu rnpu o_v n unygvat cerqvpgvba sbe C_v, naq thnenagrrf gung ng yrnfg bar bs gubfr unygvat cerqvpgvbaf vf jebat.
Va bgure jbeqf, bhg bs gur 2^x pbaprvinoyr pbzovangvbaf bs unyg naq abg unyg sbe nyy gur C_v, nytbevguz N vf nyjnlf noyr gb ehyr bhg ng yrnfg bar rkcyvpvg pbzovangvba. (Va gur bevtvany irefvba bs (3), nyt N nf V ervagrecergrq vg jnf nyjnlf noyr gb ehyr bhg ng yrnfg 2 bs gur 4 pbzovangvbaf, sbe vgf gjb vachg cebtenzf. Fb guvf arj nffhzcgvba vf nccneragyl zhpu jrnxre, zber fb vs x vf ynetr.)
Gura pbafvqre n cebtenz O juvpu gnxrf na vachg v sebz 1 gb x.
O jvyy sbez gur x ab-vachg cebtenzf O1 = “O jvgu vachg 1”, O2 guebhtu Ox qrsvarq fvzvyneyl. Vg jvyy gura cnff gurz gb N naq erprvir N’f bhgchg ovgf o_w, sbe nyy w va 1 guebhtu x. Vg jvyy gura ybbx ng o_v va cnegvphyne (erpnyy v vf O’f vachg), naq rvgure unyg be abg unyg fb nf gb znxr gung unygvat cerqvpgvba sbe Ov pbeerpg.
Erpnyy gung N’f cebzvfr jnf gung ng yrnfg bar bs gubfr o_v cerqvpgvbaf jnf jebat. Ohg O vf noyr gb znxr gurz nyy pbeerpg, nf qrfpevorq. Guvf pbagenqvpgf gur rkvfgrapr bs N. DRQ.
(Vs jr frg x = 0 va guvf cebbs, jr whfg trg gur bevtvany cebbs gung gurer vf ab N juvpu pna gryy jurgure na neovgenel C unygf.)
(V qrevirq guvf cebbs bs (3) “nhgbzngvpnyyl” sebz zl rneyvre cebbs hfvat OO, ol rknzvavat ubj vgf rffragvny erphefvba npghnyyl jbexrq. Gung tnir gur sbez sbe x = 2 naq jurer N ehyrf bhg 2 pbzovangvbaf, ohg guvf trarenyvmngvba gura orpnzr rnfl gb abgvpr (jurernf V unqa’g abgvprq vg va gur pbagrkg bs zl bevtvany cebbs, gubhtu fbzrguvat qrevirq sebz vg, creuncf zhpu zber pbzcyrk, bhtug gb jbex gurer gbb). V guvax bar pna bsgra qb guvf jvgu cebbsf gung hfr snpgf nobhg OO, juvpu va gung jnl vf zhpu yvxr Xbyzbtbebi pbzcyrkvgl — vg’f bsgra n cbjreshy trareny fgnaq-va sbe n zber fcrpvsvp nethzrag lbh unira’g jbexrq bhg lrg.)
Comment #48 May 15th, 2025 at 1:51 pm
Former Student #45:
> I think there’s a flaw in Bruce’s proof of (3): the algorithm A will not necessarily halt for both branches of computation from the oracle’s answer.
Hmm, you are right about that point. (I think this also invalidates the “simplified and generalized” proof I posted later, by making invalid its initial implicit step, though I haven’t yet tried to repair that proof.)
However, I think I can still make my first proof of (3) work, by a simple repair which works around that point (I didn’t yet read your proof) – [if anyone is still wanting to avoid spoilers, stop reading now!]
Since we know A halts for at least one branch of what the oracle might say, we can simulate both branches in parallel until that happens, and then we know one of the two possibilities I was hoping to get from it. (Though I didn’t end up needing this ability in what follows.) We also certainly know the oracle query it made, so we can compute that entire binary tree (from my first claimed proof of (3)) in terms of programs it asks about. We just don’t know how to flow the general case of an oracle answer down the tree, to get two specific possibilities for the halting or not of all the leaf programs.
But if we ever find out the true oracle answer to the top program’s halting, *then* we can safely flow down at least one step, and now we know the true answers to its two subtree-tops… and this works recursively, so eventually we know everything in that tree.
So we can still construct two programs, one which assumes the top program halted, and one which assumes the top program did not halt. Of these two programs, the one which makes the correct assumption will definitely finish computing the halting status of all the leaf programs. So it will still work exactly as described.
The other one, which assumed wrongly, may not halt, but that doesn’t matter. QED.
Now I will read your proof of (3)… I skimmed it but did not immediately fully understand it. I’ll study it more carefully later. I agree that it seems different from both of my claimed proofs (whether it might be “secretly related” to my second one, I should not yet speculate on).
But it is “close enough in spirit” that I will stop using rot13 for anything new.
Let’s see if I can repair my second claimed proof, either for its claimed more general theorem or for the original one… well, not instantly. So I retract my second one in both forms, but I still think I successfully repaired my first one.
Comment #49 May 15th, 2025 at 2:10 pm
Former Student #45:
I can also repair my “simplified claimed proof of (3)” (just for the original theorem (3) so far).
Effectively, we just make the same “binary tree”, but we only put B0 and B1 into it. Here, program Bi simulates A(B0,B1), and guesses that the oracle answer was i, uses that guess to determine which of B1 and B0 halts, runs (simulates) all the programs in that two-element set that it thinks halt, then halts.
At least one of the two programs “guessed right”. That one gets right answers about B0 and B1 halting, so it does eventually halt. But it ran itself to completion — contradiction.
(Due to your point that I missed, the other one might never halt, but that doesn’t matter.) QED.
Comment #50 May 15th, 2025 at 3:05 pm
@Bruce this is a cool generalization to (2)! But I disagree that question (3) can be reduced to it. In particular, if you have a machine that makes a single query to the binary halting oracle, then returns an answer, it is not obviously always possible to simulate both branches of it and return the possibilities (or eliminate any possible answers). What if one of the branches of computation runs forever? That would not contradict the correctness of the machine if it would not halt for the case when oracle would give wrong answer. But as simulator, I cannot distinguish between the cases:
1)It doesn’t halt in that branch, or 2) merely taking a long time to calculate an answer that I don’t know.
Comment #51 May 15th, 2025 at 3:17 pm
Former Student #45:
I now read and understand your proof of (3). I think it’s correct (and it’s simpler than it looked to me initially). If I understand it correctly, you don’t even need to assume (in case 2) that M is the *smallest* machine with its special property — any M with that property should do. Do you agree?
As for this point of yours –
> This is where the runtime bound helps: it doesn’t have to be polynomial, but if we know some computable function which is a bound on the runtime of A, we can simulate both branches of A to provide possible answers.
I had forgotten to read that part of your answer before my reply, since I had to study that flaw as soon as I understood it! I agree with this point in isolation, but I’d still enjoy seeing your proof of (2) on its own. As of right now, I don’t know any proof of (2) besides one of our proofs of (3).
> Without a runtime bound on A, it is still possible to reduce BB(a) to just one query to the halting oracle. But I didn’t understand how the advice bit can be encoded into the input to the BB() calculator, so maybe your solution can be fixed? Feel free to explain that part better.
Hopefully what I said since then has explained that. But in case not — this is just the standard way to provide advice to any program relevant to BB(n) — combine a program of length n1 with an encoding of advice of length n2, into one program with length n1 + n2 (consisting of the length-n1 program taking as input the length-n2 advice). Of course n1 + n2 needs to be at most n, for this to work. The key point is that you don’t have to know the advice yourself, to do this — you just know that among all those programs of length up to n, the one with the correct advice will be in there somewhere. If this advice helps it run long enough and then halt, it helps to provide a lower bound on BB(n), regardless of what the wrong-advice programs do (and in particular, whether or not they halt).
(This is formally the same as your proof’s saying “here are two programs, I don’t know which one of them works, but I know at least one of them does”.)
Comment #52 May 16th, 2025 at 12:11 am
Former Student #50:
I am not sure which comment of mine you are referring to here.
I never proved (2) directly, and I don’t think I did anything which I would call a “generalization of (2)”.
I proved (3) incorrectly (in two ways), read your initial explanation of the same flaw you just re-described here, acknowledged it, and (I think) repaired my proof to work around it (in both a BB-using and a simplified form).
Since some of our messages have “crossed in the mail”, I am not sure if you wrote your comment #50 before you saw my repaired proofs, or if this means you don’t think I repaired them properly. (It will help me if you clarify the comment number you are replying to.)
BTW, thanks again for posting and discussing these puzzles and proofs; I find it a really fun and interesting topic.
Comment #53 May 16th, 2025 at 12:02 pm
Only number 3?
I’d rank this as perhaps the number one blog of any sort on the Internet, let alone quantum computing.
In fact I think it is one of the greatest science resources ever created (and I said so in the acknowledgments of my PhD thesis). I still mean it.
Thanks again!
Comment #54 May 16th, 2025 at 4:08 pm
@Bruce 46: Yes I agree. Also the oracle doesn’t have to be for SAT or TQBF, an arbitrary one bit advise is not able to empower poly time algorithm to solve either SAT or TQBF.
@Bruce 47: I already responded in my post #50, your comments at #48 and #49 weren’t posted when I wrote #50.
@Bruce 48: The simplified proof still works for polynomial time (or algorithms with runtime bounded by computable function).
About the repaired proof using BB function: I think it works. The fact you are using about BB computation is that: For some fixed n, if a program can output BB(n), then it must have length >= n, which is obvious. So you have constructed two programs of length ~ O(log(n)) which iterates over all programs of length n, constructs a tree of queries by simulating A, assuming certain answer to the oracle query at the top level, and then resolve the halting status of each program at the leaf level recursively: and then calculate BB(n). Yes this whole idea is pretty nice, and it works I think.
@Bruce 49: Yes this is good, also I think it proves a cool generalization: To determine the halting status of n turing machines, Theta(log n) bits of advice is needed. This is necessary and sufficient, because if an oracle says which is the latest halting program among n possibilities, that’s enough to determine halting status of all.
@ Bruce 51: Yes don’t need M to be smallest at all, it just helps me think that M is a constant with the property. My first proof of 2 was that:
1. modify alg A to return two possibilities instead of making query
2. simulating A on (m1, m2) must reveal the truth value of either halt(m1) or halt(m2) or [halt(m1) xor halt(m2)].
3. For any pair m1, m2: let f(m1, m2) = 1 iff halt(m1) is revealed by A.
4. g(m1) = 1 iff there exist m2 for which f(m1, m2) = 1.
5. Case 1, g(m) = 1 for all m, then we can solve halting problem for m by queries A(m, m2) for each m2.
6. Case 2, exist M with g(M) = 0, then: hardcode the status of halt(M). solve halting problem for m by querying A(M, m). Since the two possibilities of A don’t reveal halt(M), it must reveal either halt(m) or halt(M) xor halt(m) which is sufficient to determine halt(m).
> As of right now, I don’t know any proof of (2) besides one of our proofs of (3).
Your unrepaired proof of (3) was good enough for (2)!
@ Bruce 52: Yes my comment #50 was in response to #47 without having seen #48-#49. you proof in #47 works for (2) but not (3) without repair, which is what I tried to say. Yes it was fun discussion!
Comment #55 May 16th, 2025 at 6:08 pm
Former Student #43
I think the solution in #45 is correct (and I find it pretty clever).
Now what about a slightly more difficult modification:
4) Does there exist an algorithm that can solve two instances of the halting problem by making one query to an oracle for any fixed decision problem?
It seems that the crux of the puzzle (3) was that the oracle can only output one of two answers but the two instances of the halting problem has four possibilities (but see (4′) below).
However, the proof exploited the fact that it was the halting problem and one of the oracle’s possible answers can be verified (i.e., if the oracle answer is 1 then we can safely run the TM until it terminates and then know that the answer is 1). I guess this works with any recursively enumerable language. So, what if the oracle is for some language that is not recursively enumerable (and whose complement also is not recursively enumerable)?
Btw there’s another modification that can be answered in positive:
4′) Does there exist an algorithm that can solve two instances of the halting problem by making one query to an oracle for any fixed ternary problem (i.e., the oracle’s output is in {0, 1, 2})?
Comment #56 May 17th, 2025 at 11:29 am
Former Student #50:
(btw, #50 is your last posted message as I write this, so I might be ignorant of some reply of yours still in moderation -)
In case you think that my simplified proof in comment #49 is still subject to the issue you identified (correctly) as a flaw in my original proof, here is why it is not – the “flaw” in this context amounts to the fact that for whichever Bi guesses wrong about the oracle answer, not only might it fail to halt due to trusting wrong answers from A, it might fail to halt merely because its simulation of the last part of A fails to halt, so it never gets those unreliable answers in the first place.
But we already knew it might never halt (due to trusting unreliable halting predictions) — having one more reason for maybe not halting, doesn’t affect the proof. The key “fix” that I added, after you reminded me about this point, is to ensure that each Bi only tries to simulate one branch of A’s later running — the branch in which A got answer i from the oracle. If it tried to simulate both branches and then use only A’s outputs from branch i, the flaw would indeed break my proof. But it only simulates branch i. Thus we can still reason that for the Bi in which i is the correct guess, its simulation of A halts, and it also gets correct information from A. It then correctly simulates B0 iff B0 halts, and B1 iff B1 halts, so it always halts, so A (being correct) told it it could safely simulate itself and it did, contradiction.
Comment #57 May 17th, 2025 at 6:09 pm
Former Student, Bruce, K, etc: Thanks, those are really nice problems (and solutions)! Sorry I was traveling the past few days and wasn’t able to keep up with the discussion.
Comment #58 May 17th, 2025 at 11:45 pm
Scott, Bruce, Physics student
It is discussions like these that make this blog the indisputable No 1 QC (even TCS, Computational Physics). Notwithstanding the rankings provided by arbitrary third party entities, it is always the community that makes the blog engaging and worth visiting frequently.
Thanks for all the effort that goes into this, which is mostly underappreciated.
Scott, hope to see more AI posts here, as you have been quite active on the podcast medium on this topic 🙂
Comment #59 May 18th, 2025 at 12:34 am
Former Student #54: thanks very much for this comprehensive answer to everything I had asked (and a bit more)!
(And Scott #57, no worries! I more or less agree with Chris #53 about the importance and uniqueness of your blog! And not only as a science resource, but for its special community of commenters.)
Here is a new puzzle inspired by this discussion.
Background facts (stated in part by #54, and K #55, also well-known in BB community):
Given n arbitrary programs, log n bits of advice suffice to allow you to determine exactly which ones halt.
One way is for that advice to identify which one halts last. Then you can simulate them all in parallel until that one halts, and then you know the whole set.
Another way is for that advice to tell you how many of them halt. Then you can simulate them all in parallel until that number of them have halted.
(In both cases, if the advice is wrong, you risk running forever when trying to use it. Both of these methods have a worst-case runtime of around the sum of the runtimes of the halting programs.)
Puzzle: suppose you are given n programs and you don’t need to know exactly which ones halt. But you are also given that (log n)-bit advice string in one of those two forms (which one halts last, or how many ever halt), and you want to know it in the other form. (You are promised it is correct.)
Is there any always-faster way than the simulations outlined, to convert between those two forms of that advice? (This is really two puzzles, one for each direction of desired conversion. The puzzle is to prove your answer, of course. Even a partial result is interesting, eg proving some lower bound on the runtime in the worst case. I just made it up — I don’t yet know the answer or how hard it is. I have “the obvious guess”, but not yet any lead to a proof!)
Comment #60 May 18th, 2025 at 12:54 am
Correction to my puzzle – it is log(n+1)-bit advice, because maybe none of them halt. In one format it either says “none of them halt” or “this one halts last”. In the other format it says “this many halt” which is from 0 to n inclusive.
If n=1 the conversion is trivial. Even if n=2, with three possible advice strings, it might be hard (and is uncomputable without promises), and I don’t know anything unobvious about it (except I think I once either heard or proved that to sort two programs into halting order (if neither halts you can answer either way) is both uncomputable, but insufficient to compute halting from, ie it is “halting-intermediate”).
Comment #61 May 18th, 2025 at 5:29 am
Hi Scott, do you have any opinion on the recent work comparing “Digitized counter diabatic quantum optimisation” to simulated annealing? https://arxiv.org/pdf/2405.13898
Comment #62 May 18th, 2025 at 8:29 am
Hi Scott,
I have a question for you, but I’m not an expert in any of this, so please be gentle with me.
Are you familiar with the Time Protection Conjecture in theoretical physics? It states that, whatever the true theory of quantum gravity is, it will disallow closed timelike curves.
I’m curious whether computational complexity could be a path to proving this conjecture.
Namely, if we prove that PSPACE is strictly larger than BQP, that means there exists some problem in PSPACE which cannot be solved in polynomial time by any unitary quantum mechanical system (imprecise language here, but as far as I understand quantum circuits are general enough to encompass all sequences of unitary transformations on a finite dimensional state space), but can be solved in polynomial time using CTCs. This shows that any quantum gravity theory which lies within unitary quantum mechanics disallows CTCs, right?
The possibility would still remain that CTCs are possible in a quantum gravity theory that is non-unitary or even nonlinear (along the lines of what Roger Penrose has talked about).
Does this idea make any sense? Or is it too vague?
Comment #63 May 18th, 2025 at 3:27 pm
Addendum: on the halting-advice-format-conversion puzzle, I made enough partial progress to confirm it is fun and interesting to think about. I won’t post any of that progress for now (it’s surely either well-known or obvious to any experts who care).
About my statement
> I think I once either heard or proved that to sort two programs into halting order (if neither halts you can answer either way) is both uncomputable, but insufficient to compute halting from, ie it is “halting-intermediate”
I recalled this well enough to be pretty sure I made it up rather than heard it, since it is partly wrong!
To clarify this “halting-order problem”: given two programs: if they both halt, say which one halts first; if exactly one halts, say that one; if neither halts, answering with either program is correct. (But the single-bit answer is only to point to one of the programs, not to give away anything about how many of them halt.)
That problem as stated is technically only a “family of problems”, since the “guessing part” (arbitrary answer if neither one halts) has to be done in some specific way, by any specific oracle which solves it.
All of them are uncomputable, by a proof similar to some of those we have discussed. But I retract my claim that any of them are “strictly easier than halting” — I don’t know the answer to that, but the proof sketch I recalled now seems wrong. It might be repairable, but I have not succeeded in doing that.
If the guessing is done carelessly (for example by lexical comparison of program text when both programs don’t halt), then the resulting oracle easily encodes the halting problem! (A fun exercise.) In fact, for an arbitrary oracle O, there is some oracle in this family which encodes O within its guesses, in a trivial way which is fast to extract. It is not too hard to create specific guess-answers which make it provably equivalent to halting (in fact, that lexical version above will do), rather than stronger or incomparable. But I didn’t yet either succeed in, or rule out, creating specific guesses that make it weaker.
Comment #64 May 18th, 2025 at 6:16 pm
(At the time of this writing, comment #57 was the last one).
Bruce Smith #48 – I read and understood the proof with BB(n) and I think it’s neat and with that fix it’s now correct.
Bruce Smith #49 – I somehow do not follow the proof as is written there. Namely:
1) What is the supposed “contradiction”? When did we assume that neither of B0 and B1 would halt?
2) If Bi is going to halt anyways why even bother simulating the subset of programs from {B0, B1} which it thinks will halt?
As I understand, you meant to write that Bi assumes that the oracle answer was i and then calculates what A says that Bi would do, and then does the opposite (i.e. Bi halts iff A said that Bi wouldn’t halt (assuming oracle answer i)). That would lead to this simplification and generalization (which also answers my question (4)):
Theorem. If a program A determines any non-trivial property for each of the k functions calculated by k programs using only one query to an oracle, then the oracle must have at least k+1 different possible answers.
Proof. Otherwise, we feed to A the programs B1, …, Bk where each Bi simulates A(B1, …, Bk) assuming that the oracle gave the i-th possible answer and then does something different of what A(B1, …, Bk) predicted that Bi would do. One of these programs guessed correctly and therefore breaks A’s answer.
Comment #65 May 19th, 2025 at 5:52 pm
@ Scott #57: No worries, I love the blog-posts and (majority of) the comment section alike. My intention of posting the questions was to get response from anybody who wants to discuss. It’s obvious that you (or anyone) could be busy at any given time to keep up with the discussion. Don’t have an expectation for replies from you specifically, but would love that anyway! I just hope you don’t mind us hijacking the comment section for (somewhat off-topic to the post but not the blog in general?) discussion that we find very interesting?!
@K 64: (last reply at time of writing): Yes that’s how I interpreted 49, and that was the proof in my mind how to show the exact theorem you mentioned. It’s quite cool.
Also, I realized some interesting philosophical point, I know the value of BB(n) function is unprovable beyond some finite value of n. However, if some alien species or God gave us the ~10kb string representation of the BB(1000) champion that we can trust, then we can settle in finite (technically, but not really!) time the knowability of any math statement that can be encoded with < 100 characters, such as P vs NP.
Proof:
1. Construct a turing machine that will that will loop over all possible proofs of P = NP, and halts if it finds a correct one. If we can encode the machine with < 1000, then we know its halting status from the 10kb advice. Hence we know if P = NP is provable (with any finite proof, not just proofs of length BB(1000) etc.)
2. Construct a turing machine that will loop over all possible proofs of P != NP, and halts if it finds a correct one. So we will know it P != NP is provable.
After this, it is still possible that neither P = NP nor P != NP is provable (with any finite proof) within our axiomatic system. I would say we have still settled the P vs. NP question at this point, since we know that no finite proof or disproof works. The only way P = NP may be true in this case is that some finite algorithm works in polynomial time, but it's not provable that it works.
@Scott I know that it has been shown that knowing BB(27) will settle goldbach conjecture. Is anything known about k such that knowing BB(k) will settle P vs NP in the above sense I mentioned above?
@ Bruce #59: For the new puzzle, let's simplify for n = 2 and a special case: If we have two TMs with promise that exactly one of them halts, how long do I need to determine which one halts? Suppose we have an algorithm A to do this.
We can reduce this to the time hierarchy theorem, that says if an arbitrary machine computes a 0 or 1 answer in f(n) time from n size input, it's not possible to simulate it substantially faster in all inputs and all such machines. Given an instance of the arbitrary time hierarchy problem: (a machine description, an input n), we can modify the machine description to embed the input inside of it, and create a version that halts if answer is zero, otherwise run forever, and another version that does the opposite of it. Feed these two versions to the promised halting solver to determine which one halts, and get the answer faster than f(n), which is a contradiction.
For the opposite direction, simplified to n = 2 and special case: Given two TMs A, B with promise that A halts, and B either runs forever or halts faster than A, how long do I need to determine which is the case?
Again, reduce this to the arbitrary machine simulation problem. embed the input inside the machine. Let A just halt just after the runtime bound, doesn't care about the output. Let B halt if and only if the answer is 0, and run forever otherwise. Feeding A and B to the promised halting solver would allow us to know the output 0 or 1 for this arbitrary machine sooner than its runtime, which is a contradiction.
Comment #66 May 19th, 2025 at 6:30 pm
I realize that Goldbach is a Pi-1 statement, but P vs NP is a Pi-2 statement. So goldbach is equivalent to “for all x f(x)” where f is decidable function. But P vs NP is equivalent to “for all x, there exist y, f(x, y)” where f is a decidable function. One way to do that is that:
for all programs p and integer k, does there exist some input u of SAT, such that program p doesn’t answer correctly the output of SAT(u) within |u|^k steps?
So truth of goldbach can be directly reduced to a halting status of a TM. But same is not true for P vs NP since it is not Pi-1. But the trinary question of probability or disprovability (or independence) of P vs NP is always a Pi-1 statement because of what I showed in my previous comment.
So 40 byte trusted advice can reduce the truth-value of Goldbach to finite computation. How much trusted advice can reduce the provability-status of P vs NP to finite computation? That is my concrete question.
Comment #67 May 20th, 2025 at 12:22 pm
K #64 and Former Student #65 – Thanks for your replies! With respect, I think you both misunderstood my argument in #49. Rereading it, I was terse, but did not misstate anything. The contradiction is not that one machine both does and doesn’t halt. It is that one machine, which halts, does so after a run which includes fully simulating itself running and halting (as well as other things). This is not a contradiction in halting status, but in runtime, since no finite runtime can be larger than itself.
K #64’s theorem about k programs and one oracle – I agree, and I think it’s a nice generalization of this kind of thing. And (I think you already agree, I just wanted to make it explicit) it doesn’t even require that A halts unless it gets the correct answer from its oracle query.
FS #65 replying to me #59 about the “new puzzle” (let’s call it the “halting-advice format conversion puzzle”) – I am eager to read and reply, but I’ll have to save that for later.
Comment #68 May 20th, 2025 at 12:40 pm
Former Student #66:
(this one is much faster to reply to)
> So 40 byte trusted advice can reduce the truth-value of Goldbach to finite computation. How much trusted advice can reduce the provability-status of P vs NP to finite computation? That is my concrete question.
I am not sure exactly what you mean, but if you receive trusted advice which means to you that P vs NP is decidable, that has already reduced its truth to a finite computation (search for the proof of one or the other way of deciding it). In a sense that is 1-bit advice (ie a 1-bit answer from a “decidability oracle”, relative to a specific theory to be used in the proofs).
Comment #69 May 20th, 2025 at 3:59 pm
@Bruce#68 yes it doesn’t make sense to talk about decidability of a single binary question in the presence of trusted advice, since the answer can be part of that advice.
What I am really trying to say is that if we had 10kb trusted advice (my best guess), it allows not just deciding provability of not just p vs np, but any math question of a certain length. But if we want to focus on p vs np, then we can obviously frame it as what is an upper bound on the number of states of a machine whose halting status is equivalent to the provability of p vs np? (In the same way that it has been been shown that goldbach conjecture is equivalent to the halting status of a 27 state machine).
It’s not really a question that’s possible for us to answer. But it’s interesting to me that there exists some finite, and probably pretty small advice that would allow creating a decider for any math question below some complexity threshold
Comment #70 May 20th, 2025 at 5:44 pm
Prasanna # 58:
No. Don’t. Don’t write like that. No. Not ever.
If at all I’ve in the past hated Scott, one of the most prominent reasons has been that which you now illustrate. [Yes, sometimes I also errr… expressed my errr… displeasure to Scott. And, also to his PhD Guide’s email ID at Berkeley. That’s just me, you know…. ]
It doesn’t matter [and it didn’t, not to me, not even back then] that JPBTIs and their equivalents in the USA would hate me. Which, they did. [Rather, they *started* it.]
Still, a piece of an advice. Being given, simply because Scott is some 20 years junior to me, and you obviously are even more junior:
No. Don’t do that.
Be sincere in your compliments. Utterly sincere. I don’t care if you rant (and that too, within reason). But, when you do compliment, follow these two rules: (i) Be realistic (i.e. Reality and Reason oriented), and thereby, easy, (ii) be sincere.
Or, if it helps you: Remember: “We must be reasonable in our request.” (It’s a highly diluted form of what Aristotle, or at least the Aristotelian tradition, once said. It was detectable right in my life-time, especially at Harvard (which I never attended), but neither at Berkeley nor at any other US / UK university — Newton’s Cambridge and Oxford included.)
If you follow these two rules, you would *actually* be more effective than you ever could be, following your that strategy from which your above example came.
Yes, you will be poor. So what?
You would be more “effective” (or, to use a non-mathematical and more philosophic term: “efficacious”). And, eventually, more happy. Even CS JPBTIs from IIT Bombay like Dr. Navin Kabra (PhD, CS, Maryland?) would want to “follow” you “up.” Perhaps, by hacking your laptop / ‘net. Or, joining the WhatsApp groups of those who do. (End-to-end encrypted, alright?) Why?
Answer: “Efficacious-ness.” Yours. Your own.
And, frankly, the way you write, and get away with it, even in your age, I don’t think, really speaking, that you will be poor.
Best to you both (Scott and you),
–Ajit
Comment #71 May 20th, 2025 at 7:47 pm
Which side has truth on it’s side: Christianity-Islam-Atheism or Judaism-Hinduism-Jainism?
Comment #72 May 21st, 2025 at 5:40 am
@Scott: Congratulations, and keep up the good work.
Regarding the ‘cranky’ emails, say someone is reasonably bright, maybe even holding an advanced degree in a hard science, but not exactly a specialist in the field. S(he) has too much time at hand and chooses to get involved in things like p=NP, quantum gravity, you name it. Obviously, first attempts will contain embarrassing beginner’s errors, fallacies, etc. So whom does s(he) talk to now not being at a university anymore? There must be some life between vixra and PRL I hope…
Comment #73 May 21st, 2025 at 1:27 pm
onlooker #71: That must be one of the strangest questions in the long, strange history of this comment section!
Is the common ground that Christianity, Islam, and atheism are all (at least in some versions) proselytizing movements whose ideal is to convert the entire world, whereas Judaism, Hinduism, and Jainism aren’t?
Comment #74 May 21st, 2025 at 1:47 pm
Former Student #69:
When I replied in my #68 I hadn’t yet read your #65. I think your question in #65 was clear:
> @Scott I know that it has been shown that knowing BB(27) will settle goldbach conjecture. Is anything known about k such that knowing BB(k) will settle P vs NP in the sense I mentioned above?
I would like to hear Scott’s answer too! But my guess is, it will essentially just be the sum of whatever k for which it is known BB(k)’s value can’t be proven by whatever formal theory you have in mind (e.g. PA or ZF) (which Scott has reported on but I haven’t kept track of, but iirc it was less than 1000), plus whatever additional program-space it would take to encode a recognizer for some statement equivalent to P = NP in that same theory; and that the additional program-space will be “small by comparison”.
That is because the existing limit on this k (for which BB(k) isn’t provable) will have been based on a machine which searches for a proof of 0=1 (or any known false statement) in that theory. To change it into the machine you want, we just have to change its “recognizer for a proof of 0=1” into a “recognizer for a proof of P = NP or its negation” (or make two machines, one for each of those statements, if that saves a byte or two).
I’m guessing the main complexity will be in “recognize a correct proof” vs “recognize that the final theorem in that proof is statement S” (for this level of complexity of S), thus I guess this new k will be “not that much higher than the old one”. (But since I have not actually formalized either part, I could be wildly off about their relative complexity.)
(Personally that’s all I need to know about it! The real interest to me is in the proof technique, and the fact that k is “not huge”, rather than its specific value, which depends greatly on details of the TM model you are using, and even on details of how you choose and express the formal theory you use. For example, it is not obvious to me that choosing a more powerful theory, say ZF over PA, could not make the value of k go down, if the more powerful theory was simpler to express in your TM model. I am not aware of that being a contradiction (unless your proof assumes/uses the consistency of one of the theories involved, or at least their relative consistency, in order to conclude that the same machine can serve both purposes). I would be happy to be proven wrong if someone spelled out that proof!)
(Chaitin has worked out the analogous number to this k for his binary lisp machine model of TM, for the closely related problem of how high a Kolmogorov complexity can be proven for a specific string, and the result can be looked up, and was somewhere in the range 2000-5000 bits iirc.)
Comment #75 May 21st, 2025 at 1:56 pm
Prasanna #58: Thanks! (Re my small implicit share in your comment.)
BTW I have no reason to doubt your comment’s sincerity, and I agree with its praise of this blog. It provides a forum which many of us find uniquely valuable. If not everyone does, that neither surprises me nor worries me.
Comment #76 May 21st, 2025 at 2:03 pm
Former Student #65, re your comments on the “halting-advice format conversion puzzle”:
Your first comment sketches a proof that, for converting from “advice k = how many halt out of n” to “which one halts last out of those n” in the special case n = 2 and k = 1, an always-correct alg A can’t in general do it much faster than the runtime of the one which halts.
I agree with that result and proof.
(And there is an obvious alg for A which will do this in time no worse than a bit more than twice that runtime. So it’s solved to within a factor of around 2 * log(runtime). I suspect any tightening of that gap is unknown and very hard, either for the “2 vs 1” part (can you save time in a parallel sim of two unrelated machines) or the “log runtime” part (based on known sim overhead, though this may depend on your model of computation).)
I agree with your other result and proof too (on that puzzle), and with it being the same special case you implied. (In both cases, I did not actually check your proof details, only that they plausibly sound like a proof I’m sure exists which I would agree with. I just mean that if you switched 1 and 0 somewhere, etc, I did not try to check for that.)
That result has the same gap between lower and upper bound, since the obvious correct alg for A (the algorithm, not the first input program) also has to simulate the two input programs A and B in parallel, to use A’s runtime as a limit on B’s possible runtime. (So it can run not much longer than twice the runtime of whichever of them is faster.)
Comment #77 May 21st, 2025 at 7:08 pm
?? https://www.yahoo.com/news/china-debuts-world-first-quantum-181931993.html
Comment #78 May 22nd, 2025 at 10:25 am
Scott (feel free not to post),
I keep coming back here expecting a lengthy post on Columbia/Peter Woit/etc. Perhaps it’s best not to have more discussion on this, I don’t know.
Regardless, I just wanted to say that I think you’ve conducted yourself a lot better than Peter over in his comments section, both in explaining your arguments and in terms of arguing in good faith.
Like anyone, I’m biased on this topic, but if I (do my best to) ignore the topic being discussed, it’s been disappointing to see Peter refuse to engage honestly, to deflect with accusations of lunacy/paranoia/delusion/etc., claim umbrage when eventually responded to in kind (“antisemitic piece of shit”, oh well), and report that most of the other comments (which, of course, he won’t publish) support his claims of your idiocy, and so on.
On political topics, I agree with perhaps about 70% of what you say. But even when I occasionally complain about you to my wife (“Texas Scott has really got it fucking wrong”), I never suspect that you are arguing in bad faith, or out of lunacy.
Anyway, take it easy.
Comment #79 May 22nd, 2025 at 11:31 am
Scott: I’m currently terrified about AI alignment, because of this piece http://www.ai-2027.com that I read. Can you say something to put my mind at ease?
Comment #80 May 22nd, 2025 at 2:23 pm
Scott: I’ve had a horrible week, highlighted by the Eurovision incident (an Israeli singer who was a survivor of the October 7 attempted Holocaust was booed by the audience), “anti-Zionist” bullshit at my university, more horrifying details I read about what happened to the October 7 victims, and Peter Woit, a guy who I used to respect and admire, succumbing to antisemitism. I second AlgorithmsGuy here, and I have to say as an outside observer, you came across as measured and logical in your debate with Peter, while Peter came across as angry, irrational and unhinged, and honestly, motivated by some kind of hatred, although I won’t go so far as to say he hates Jews. I’m sorry for the comment I submitted earlier about him, which went too far.
That exchange was extremely upsetting for me, and it hurt to see a hero of mine behave this way, with so much contempt for my people and for my fears. It would be very helpful if you could post about this exchange on your blog, so we can “debrief” and process what happened, thanks.
Comment #81 May 22nd, 2025 at 3:59 pm
Could you please edit my comment to also say the killing of the two Israeli embassy workers killed in DC, to the list of depressing things happening this week?
Comment #82 May 22nd, 2025 at 8:03 pm
Like others I would be happy to have a place to vet my frustration and disappointment with Peter Woit’s behaviour. At first he let me post comments on his website, however, he did filter out my last two comments which criticized him. Just to be clear I am very angry with Israel’s government and its voters. Nevertheless, this does not mean the Palestinians are just innocent victims. I don’t agree with everything you say, but you definitely conducted yourself in a calm and rational way and in good faith. His position seems that there is nothing wrong in Columbia and worse that anyone who feel differently is essentially crazy. This drives me into considering that maybe closing one or two Ivy league institutions might remind other institutions that they serve the public and not the other way around. I am not talking just about antisemitism, but on many other issues like social mobility and scientific corruption.
Comment #83 May 22nd, 2025 at 8:55 pm
AlgorithmsGuy and Julian: Sorry, I’m pretty much a pure ball of fear and depression right now — a large fraction of the world (including many parts that I’d thought were ok) wants me dead, for being a Jew who favors continued Jewish survival, and there’s no argument on explanation that I can possibly give to change its mind. And I don’t know whether blogging about it would make things better or worse. In any case no one else should be coming to me for advice right now. Just try to stay safe.
Comment #84 May 23rd, 2025 at 1:23 am
What’s your take on the revokation of Harvard student’s legal immigration status?
Comment #85 May 23rd, 2025 at 4:19 am
@Scott #83
>> a pure ball of fear
I think there are two facts you need to consider:
i) >> a large fraction of the world wants me dead
Unfortunately, nothing changed imho, just that the hate is now visible thanks to social media and money from Qatar etc. supporting the idiots.
But it is better to be aware of the truth than to live in a comfortable illusion.
ii) The security situation of Israel is objectively much better than it was a few years ago.
The Sinwars are dead, Hamas is mostly destroyed, Hezbollah is significantly reduced and I would think that the days of the mullahs in Iran are numbered.
In other words you should consider moving to Israel and live a life mostly free of fear …
Comment #86 May 23rd, 2025 at 8:14 am
Fan #84: It’s horrible, obviously. As should go without saying, the overwhelming majority of international students at Harvard have nothing to do with antisemitic incitement, and are just having their careers destroyed for no reason. I personally know some of these students, and will try to help if I can. I hope that Harvard gets this overturned swiftly in court.
Comment #87 May 23rd, 2025 at 8:38 am
Scott #86: I’m sorry, I don’t want to do this because I know you’re feeling so bad, but I have to disagree with you there. Where did you get this fact, that the “overwhelming majority of international students have nothing to do with antisemitic incitement?” I went to uni in the States before my current program, and there were many international students from Africa and the Middle East, many Muslim international students, and the vast majority were vocally anti-Zionist and “anti-colonialist.” In fact, they were obnoxiously opposed to Western societies and what they perceived as “colonialism,” which somehow included Israel. They were very anti-American to boot, despite being invited to study there. I was there before Oct 7, but I have no doubt that they’re currently anti-Israel. Of course there were exceptions. Most Iranians were actually fine, surprisingly enough. As for students from East Asia, it was a mixed bag, some of them woke, some of them not. But most of the Muslim internationals were obnoxious anri-Zionist, anti-American, far-left radicals.
As President, you have to make tough decisions. Yes, there will be international students who AREN’T anti-Zionist assholes whose careers are ruined by this, and I feel terrible about that, I’m not a monster. It’s tragic. But this is a necessary evil to expel anti-Zionist extremism from our top universities, and make them a safe place for Jews and Israelis. Making universities a safe place for Jews should be the top priority, and unfortunately the president has limited levers to pull to make the universities comply. Some innocent students’ careers will be derailed, but I blame Harvard for not capitulating. If they’re so desperate to remain a safe space for anti-Zionist radicals, they will suffer for it, as Trump is showing. I hope they will comply soon so that Jewish students will be safe AND innocent internationals can continue their careers.
Comment #88 May 23rd, 2025 at 9:10 am
Julian #87: What makes you an expert on what levers the US federal government does or doesn’t have? They can revoke the visas specifically of those students who engage in prohibited incitement, and they can go through a Title VI process to make funding contingent on universities enforcing their time, place, and manner rules.
Also, in American STEM, the majority of international students come from China and India. It wouldn’t surprise me at all if those students were involved in anti-Israel activism at a lower rate than native-born students. They’re just pure collateral damage—and lots of groundbreaking science involving those students will be collateral damage as well.
Comment #89 May 23rd, 2025 at 9:30 am
Scott #88
> They can revoke the visas specifically of those students who engage in prohibited incitement
They could, if Harvard would share the relevant data, which according to the Trump administration it refuses to do. See the full letter here, in case you missed it:
https://x.com/charliekirk11/status/1925625458487369729
Comment #90 May 23rd, 2025 at 9:33 am
Hi Scott,
I’m far from an expert on constitutional law, but I expect that’s also true of most of your readers who are commenting on the events at Columbia and Harvard. That said, I’ve spent a lot of time trying to understand what the Trump administration is doing here and the options available to them, and I think I have some understanding of it.
Singling out students for deportation based on anti-Zionist attitudes is constitutionally problematic. It may or may not contradict first amendment jurisprudence. Hopefully the Supreme Court will rule on this soon. At least one federal judge ruled against it. Personally, as much as I despise anti-Zionist students, I am actually opposed to deporting students because of anti-Zionist speech. I’m against deporting the Tufts student for example or the Columbia kid. It’s a matter of principle for me. The government should not be allowed to revoke visas based on speech.
On the other hand, blocking ALL international students at Harvard seems to fall squarely under the President’s constitutionally enumerated powers.
Unfortunately, the threat to withold funding from Harvard was not enough to force the university to commit to reasonable concessions to make them a safe environment for Israelis and Zionists. The administration had to pull additional levers to force their hand in this matter.It’s unfortunate that this is one of the remaining levers they had available to them.
I do not think enforcing “time, place and manner” rules surrounding protests is sufficient to make universities a safe environment for Jews and Zionists. For instance, this would still permit students to say “go back to Auschwitz!” if they weren’t disrupting classes or using a bullhorn, for example. Also, there’s an overwhelming anti-Israel bias among humanities faculty that makes the university an uncomfortable environment for Jews. It’s hard to address this kind of thing through Title VI. It’s hard to address the constant, for lack of a better word, “microaggressions” against Zionist students in the dorms and in student clubs through the existing Title VI infrastructure. I think the Trump administration’s suggested reforms are reasonable in this regard.
As I said, the American uni I attended was full of Muslim internationals. Most of them were not in STEM. But they were there, and they hated Jews, and continue to hate Jews.
Here’s a good analogy. In World War 2 we firebombed Dresden. We burned alive countless innocent civilians to defeat the Nazis. Horrible collateral damage. Was it worth it to defeat the Nazis? I submit that it was. And right now, there is a nascent neo-Nazism in the guise of far-left ideology that is festering at American universities. Some collateral damage of innocents is reasonable to nip this in the bud.
Comment #91 May 23rd, 2025 at 10:12 am
Julian #87, thank you for adding a data point confirming my gut feeling that Muslim international students are often “anti-Zionist” (in Scott’s sense) and that Iranian students are actually “fine”. I also suspect that often female students coming from the very religious Islamic countries and Turkish students are “fine”. All of these groups are groups which are oppressed by religious Muslims in their own countries. It is one of the great tragedies of the Middle East that Iranians and Israelis find themselves in on two different sides. There is a lot of common between these two countries. In both there are big groups with liberal point of view (but that also hold some traditional values, like family values). Unfortunately, this is becoming true also with regard to Turkey and Israel.
Nevertheless, punishing all international students for the sins of some is not a good idea. I would rather see a more selective process of taking students and, much more importantly, of hiring staff members who has more balanced views regarding liberal values, especially in humanities.
Comment #92 May 23rd, 2025 at 11:36 am
It seems a bit silly that their lists includes such “blogs” as Nature, Fermilab, and MIT News. That’s not exactly what I imagine when I hear the word “blog”.
Comment #93 May 23rd, 2025 at 11:45 am
Julian, the problem with such navel-centric collateral damage discourse is that everyone else may well say that the microaggressions suffered by Israelis and Zionists are small collateral damage in the face of the disruption of losing all the international talent and the defunding of the university. And who is going to say otherwise when the alternative you propose is so bleak?
And I’m not American, but for presidential power to include a guy getting up in the morning and saying that the company pointed at by his finger can’t hire whoever he wants (those that for example can be hired by their competition), or that many Americans want it to be so, marries very poorly with anything resembling democracy.
Comment #94 May 23rd, 2025 at 3:07 pm
Zionism is a political direction, and criticizing it is perfectly ok like with any other political direction.
Antisemitism is hating Jews because of their religion and/or „race“ and of course to be despised and legally punished.
I fail to understand why many in this blog fail to recognize this simple distinction.
Comment #95 May 23rd, 2025 at 4:53 pm
Georg #94: In theory you’re perfectly correct (and certainly you would’ve been correct a century ago).
In practice, now that the State of Israel exists, and is the only thing standing between half the world’s Jews and their violent deaths in a massively scaled-up October 7th, to want to see Israel eradicated (i.e., to be an anti-Zionist) strongly suggests an attitude of indifference, if not outright joy, at this prospect of a second Holocaust.
Then again, as you surely know, many leading Nazis took the position that the Holocaust itself wasn’t antisemitic: there was no personal animus on their part against Jews, you see, just a political decision that the earth would be better if all Jews were gone. Perhaps many anti-Zionists today feel the same way.
Anyway, I fail to understand why so many fail to recognize these simple distinctions!
Comment #96 May 23rd, 2025 at 4:55 pm
I was not able to post this comment on “Not Even Wrong”:
The following reaction to your exchanges with Scott is informed, in part, by “The Situation at Columbia VIII”, in which 1933 Germany is invoked repeatedly, e.g.
“In 1933, anti-Communists were generally happy to see someone come in and rid their institution of the Communist problem, even if they didn’t otherwise support the new dictatorship.”
To the best of my judgement, it is the consensus of the serious historians of the period that the Communists in Germany in 1931-33 made a mistake: following Stalin’s misguided instructions, they refused to ally with with SPD (Sozialdemokratische Partei Deutschlands) to form an anti-Nazi coalition. It was this fatal error that made possible Hitler usurping power in Germany in 1933.
Comment #97 May 23rd, 2025 at 5:03 pm
When the Apartheid regime ended in South Africa, were the white Afrikaners exterminated? I’m sure in the 80s many fear-mongering Apartheid politicians predicted this! I don’t understand why dissolving the colonial regime won’t end like South Africa—a democratic, binational state with equal rights for everybody. That’s the obvious solution.
Comment #98 May 23rd, 2025 at 5:28 pm
Georg #94, there are two options either Israel is a Jewish state (the only Jewish state) and then calling to the dismantling of Israel is antisemitic. Or Israel is not a Jewish state and then calling to the dismantling of Israel is racism again Israelis. In any case, the people of Israel have a right for a state of their own as any other nation, the same as the Palestinian people.
Comment #99 May 23rd, 2025 at 5:52 pm
Emma #97: One obvious difference between the two cases is that the blacks of South Africa never tried to exterminate the whites, as the Arab nations tried to exterminate the Jews. While there was certainly violence, there was never a whole theology extolling the destruction of all whites as the highest moral good.
A second obvious difference is that the South African whites (until 1994) never offered equal rights of citizenship to those blacks who wanted to live with them, nor did they offer an independent state to those blacks who didn’t. If they had, the South African blacks would’ve accepted either offer, rather than continuing to pursue the dream of killing or expelling all whites.
A third obvious difference is that whites were never indigenous to South Africa, as Jews are indigenous to Israel.
But none of these differences will make a dent for you—you can’t reason someone out of a position that they weren’t reasoned into in the first place.
Comment #100 May 23rd, 2025 at 6:26 pm
I absolutely oppose the administration’s actions to control / punish Harvard and other academic institutions. I think all speech / protest / publication regarding israel-palestine conflict should be fair game as long as they are not actively blocking / disrupting other students in these institutions. There is heightened risk of such disruptions / violence towards specific groups as a side effect. That should be handled by exemplary harsh punishments when the illegal behavior is proven through rigorous investigation and the evidence / facts should be made public.
Rather than punishing the university as a whole without any hard evidence, why not bind them by some contract to protect the students, and allow the students to sue in specific cases where the students feel university didn’t do enough to protect them? Are there any such student (jewish or other) at hardvard who are at least claiming they have not been given justice? Why does anyone else have standing to take action?
Comment #101 May 24th, 2025 at 1:33 am
AG #96, no doubt a unified left against Hitler would have made it considerably more difficult for the Nazis (although not necessarily impossible) to seize power. But with regards to Stalin’s position there was nothing misguided in the instructions he gave the German communists. Clearly he saw Hitler as a potential ally from the get-go and certainly must have felt vindicated when he entered the Molotov–Ribbentrop Pact to invade Poland in alliance with Nazi Germany. Hitler obviously (and fortunately) did not feel the same way. I shudder to imagine what history would have looked like had he maintained that Nazi-Soviet alliance.
Comment #102 May 24th, 2025 at 3:01 am
I haven’t posted in a while since wasn’t sure I could tolerate the wild ideologically driven expectations applied to the current US political situation but I have a request-
If anyone should hear that Peter Woit has renounced his US citizenship (fat chance) will they please post that information here. I intend a small celebration party with friends.
Also the real problem to be addressed with the Ivies is not just the disruptive pro-Gaza activities on campus but that the Ivies as a rule produce students with this pro-Gaza belief set.
You can read through his recent posts and see exactly why the Ivies have become objects of derision for much of the population of the US. Arrogance (that I don’t understand in this particular case) and hatred. The ivies were entrusted to educate the elites and the elites have led us to where we are now. Where we are now is a poor educational system and overpriced medical care with poor results and too often based on fraudulent scientific results and all of the poor decisions leading to this soundly based on left wing ideology. It’s the usual attack on empiricism with assurances that there is no athletic differences between men and women and people commit crimes because of the racist criminal justice system.
I agree with a comment that you made on his blog that dialogue cannot bridge this gap so celebratory for me if in fact Peter Woit and those like him renounce their US citizenship. It would be best for everyone and I would hope that Jewish people remain in the US and continue winning Fields Medals and Nobel Prizes.
Comment #103 May 24th, 2025 at 3:10 am
As I have stated before-Ironic that the independence of the academic community was intended to protect diversity of thought but in fact it is used to protect uniformity of thought.
There may be hope now for the future queen of Belgium.
Comment #104 May 24th, 2025 at 3:29 am
Also congratulations on the recognition but you are certainly at the top of the fifty.
You seem to have mellowed over the past few years. You used to have posts with similar emotional content to Woit’s post but a more positive tone now. One of my early posts on this board was that the left always seems to eat Jewish people first (envious of high distribution of ability in the Jewish population) and I expected that would be the case in the US. Admittedly I didn’t expect it to happen so soon.
Comment #105 May 24th, 2025 at 4:46 am
Scott #95:
Zionism today promotes ethnical cleansing in the Westbank and Gaza, as outspoken by the corrupt prime minister and his right extremist partners in Israeli government.
Being against this does not mean to question the right of existence of Israel, and of course not „an attitude of indifference, if not outright joy, at this prospect of a second Holocaust.“. This insinuation of yours is outright insulting.
Comment #106 May 24th, 2025 at 5:27 am
Georg #105: Zionism means support for the existence of a Jewish national home in Israel. The word has never meant anything other than that. If you support Israel’s existence, then you too are a Zionist (welcome to the club!).
If you’re against Netanyahu, say you’re against Netanyahu. If you’re against settlements, say you’re against settlements. If you’re a liberal Zionist who favors a two-state solution, say that. (I’ll be with you on all of that.)
Even if you think Israel should end the Gaza war unilaterally and leave Hamas in power to plan its next murderous invasion of Israel … even then, you can still be a Zionist.
If you say you’re an anti-Zionist, all 90% of the world’s Jews will hear, is that you want half of us to be disarmed and defenseless against the next Holocaust. Because that’s not at all a strawman position, but one that’s wildly popular all around the world, as a couple hours on Twitter will confirm for you. I hope that makes things clear.
Comment #107 May 24th, 2025 at 7:27 am
Scott #106
“Zionism means support […]
If you’re against Netanyahu, say you’re against Netanyahu. If you’re against settlements, say you’re against settlements. If you’re a liberal Zionist who favors a two-state solution, say that. (I’ll be with you on all of that.)
[…]If you say you’re an anti-Zionist, […]
“
Thank you for once again explaining so clearly what Zionsim and anti-Zionism mean and imply.
Netanyahu now openly admits he’s using starvation of the civilian population of Gaza as a weapon, with the goal to reach Trump’s solution to have them entirely leave one way or another, by slowly and incrementally making life impossible there.
For many people, using starvation on a civilian population is a war crime, and forcing the relocation of a population out of their land is ethic cleansing.
If someone doesn’t support this, what term should he/she use to neatly express it?
Do you have any handy …ism or anti-…ism terms to call this?
Comment #108 May 24th, 2025 at 8:38 am
Scott #106
> Zionism means support for the existence of a Jewish national home in Israel. The word has never meant anything other than that.
Zionism meant support for the *establishment* of a Jewish national home in Israel. Things were touch-and-go for the first ~30 years of Israel’s existence, but for at least as long it’s been abundantly clear that Israel is here to stay. That’s why I – an Israeli Jew who spent most of the year following October 7 in uniform – don’t think of myself as a Zionist, and frankly find such self-identification a bit weird, like an American calling themselves a (slavery) abolitionist. Similarly, I think of anti-Zionists as I would of people who’d declare they’re pro-slavery: contemptible, sure, but mostly silly, because the issue was settled long ago.
Comment #109 May 24th, 2025 at 9:11 am
Vladimir #108: Thank you for your service. I’d say that you do count as a Zionist. 🙂
As you know, things look radically different in large parts of the world, where the default position is still that Israel’s establishment was a terrible mistake that can still be reversed (and if that reversal leads to a second Holocaust … well, the Jews brought it on themselves). This is the position that’s currently winning all across social media, and also winning with the youngest generation of Americans.
So, while it might seem strange to explicitly identify as a round-earther, in a world of flat-earthers what choice do you have?
Comment #110 May 24th, 2025 at 9:23 am
fred #107: If you don’t want innocents to starve to death, that just makes you an ordinary decent human being.
But can you understand the dilemma? For almost two years, Israel has allowed aid to flow to its wartime enemy. And for almost two years, Hamas has commandeered much of the aid for its own fighters, sold the rest at wild prices, and used the proceeds to remain in power and continue this horrible war. So finally Israel said: we’ll continue the aid, but the IDF needs to handle the distribution itself. Incredibly, the UN and international agencies screamed that this is totally unacceptable: the aid needs to flow to Hamas and only to Hamas, or else Israel is guilty of deliberate starvation.
What would you do in such a situation, where your normal human decency was weaponized against you by those who wanted you dead? (To be clear, I don’t regard the answer as obvious.)
Comment #111 May 24th, 2025 at 9:32 am
Scott #110
Ok, but I didn’t actually ask you to defend/justify the tactics, I know where you stand (the end justifies the means), I really simply asked you to please come up with a handy term to call one’s opposition to those *specific* war tactics of using starvation and population relocation used on the ground in Gaza by Netanyahu (again, even he doesn’t deny it).
Comment #112 May 24th, 2025 at 10:09 am
I hope that anyone reasonable would agree that one’s position for or against *every single particular/specific matter/question* of what’s going on in Israel/Gaza (e.g. the tactics used to root out Hamas) just can’t always be classified into two boxes: one called “pro-zionism” and one called “anti-zionism”, splitting humanity in exactly two camps.
It makes no sense since the same person would find himself in one box for some questions and the other box on other questions.
At the very least you’d need a third box called “Doesn’t apply”, but we’re not even given that luxury – everyone is being forced into those same zionist/anti-zionist boxes no matter what’s being considered, because it serves the “the end justifies the means” camp.
So I’m asking for a set of boxes for the specific question of using tactics of starvation and population relocation on the Gaza population.
I guess there’s always
Pro-Gideon-Chariotist/Anti-Gideon-Chariotist… but it’s really no laughing matter.
Comment #113 May 24th, 2025 at 10:12 am
Scott #109
I’m not suggesting that pointing out that “Zionism – pro or con?” is last century’s debate would be a rhetorically effective response to someone like Georg #105, never mind the tentifada people. I do think that many diaspora Jews could benefit from internalizing this mindset.
Comment #114 May 24th, 2025 at 10:19 am
fred #111: Before I can answer your question, I actually need to know your answer to mine. Recall:
What would you do in such a situation, where your normal human decency was weaponized against you by those who wanted you dead?
If you would let yourself and your whole family, including your kids (if you have any), be burned alive or shot point-blank, in order to save the wives and children of the genocidal enemy who’d set out to kill you, then your position could be called “self-immolating altruism.” If you wouldn’t, then it’s just standard liberal Zionism.
Comment #115 May 24th, 2025 at 11:15 am
For me an easy question to answer is-Which of Hamas or Israel actually has genocidal intent My selection even openly acknowledges the objective.
Comment #116 May 24th, 2025 at 4:36 pm
I’m sorry to hear you’ve been depressed from what I assume are online comments. It is a difficult thing to see open hatred against you without letting hatred consume you in turn, I know that from personal experience, seeing such things have made me a smaller, meaner person. I can’t give you any words of encouragement or advice here, hopefully this small act of sympathy will help, however little.
Comment #117 May 24th, 2025 at 6:35 pm
Max Madera:
“Navel gazing” in this case is my audacity to be afraid as a Jew after thousands of my fellow students celebrated the rape, torture, and murder of Jews just for being Jews. If you celebrated October 7 you are a piece of shit and should be sent on a slow boat back to your country, if you are from another country, and if you’re from here you should be expelled and/or lose your job.
Do you know what happened on October 7? Do you know what they did to the women and children? Do you know about the people stuck in their cars stuck in traffic leaving the music festival as the Hamas terrorists moved from car to car slaughtering them? Can you imagine their utter terror?
And the students CELEBRATED this. And the universities LET THEM, even though they’re so fucking woke about everything else, like disciplining students for being sexist or whatever.
So universities say even though you’ll be expelled for sexual harassment or whatever, or racism or whatever, it’s totally fine to celebrate the mass slaughter of Jews. Because let’s be real, they’re controlled by anti-Zionist far-left freaks who want to see me and my family dead.
And of course I’ll support Trump who is the guy saying he’s gonna keep me and my family safe from these freaks and punish the universities for what they’ve done and punish these murder-supporting bastard students.
But that’s just “naval gazing.”
You are a disgusting human being.
Comment #118 May 24th, 2025 at 7:27 pm
OMG #115
yes, intent is terribly important, but it’s nothing without actual agency.
One example:
The citizens of America, the supposed greatest “representative democracy” in the world, have no say whatsoever over the fact that their taxpayer money is being used to fund/support Operation Gideon’s Chariots (and this is true regardless of who would have won the last elections).
American voters obviously don’t have a say either into who rules Israel and ultimately decides how the generous American funding is being used.
In these matters, the agency of the American voters has basically been fully transferred to the Israeli voters.
With such a great power comes a great responsibility.
Comment #119 May 24th, 2025 at 8:33 pm
Is it fair to say that opposite of Zionism is Islamism?
Comment #120 May 24th, 2025 at 8:40 pm
fred #118: Until pretty recently, Israel was strongly supported by solid majorities of both Democratic and Republican voters. So your complaint about the electoral system boils down to saying that a loud anti-Israel minority was unable to impose its will on the majority, which … uh, no shit?
Now that a segment of Democratic voters has turned against Israel, for better or worse we see a growing faction of Democratic politicans (e.g. the Squad) turning against Israel as well.
(1 point for Occam’s Razor, 0 points for antisemitism-tinged conspiracy theories… 😀 )
Comment #121 May 24th, 2025 at 10:55 pm
Hi Scott #120
“Until pretty recently, Israel was strongly supported by solid majorities of both Democratic and Republican voters. So your complaint about the electoral system boils down to saying that a loud anti-Israel minority was unable to impose its will on the majority, which … uh, no shit?”
Indeed, “until recently” is key here, because the Netanyahu government hadn’t made it so crystal clear that the use of starvation and population relocation was part of the plan.
Just like “until recently”, lots of Americans were on Trump side, until things evolved and now many are saying “this isn’t what we signed for”.
And what’s also key is that I wrote “funding for Operation Gideon’s Chariots” very specifically.
But of course you immediately generalize and conflate this automatically with being “anti-Israel”, which is so conveniently vague that it can cover anything from being anti-Netanyahu to being for the genocide of all Jews.
It’s also striking how it’s ok for Republicans to have been bitching for years about their tax payer money being used to help Zelensky fight Russia, but once the same point is being brought up on the funding of Netanyahu’s policies, suddenly it’s a matter of being into conspiracy theories.
And since you keep making innuendos that I’m driven by antisemitic conspiracy theories (glad to see that at least something makes you laugh):
Jews don’t rule the world any more than anyone else, the rich rule the world.
Rich Israelis have their chance to influence the US, but, as we’ve seen very recently, so do rich South Africans, rich Qataris, and rich Saudis.
At least, everyone has their chance at corrupting the US representative democracy, and those corruptions from all sides is what will balance out the US policies in the end! 😀
Comment #122 May 24th, 2025 at 11:48 pm
fun #119:
Is it fair to say that opposite of Zionism is Islamism?
No, it isn’t. The opposite of Zionism is the belief that Israel should be destroyed, whether that comes from Islamism, radical leftism, Nazism, or any other ideology (even the fringe Orthodox sects that oppose modern Israel’s existence because it’s secular).
Comment #123 May 25th, 2025 at 12:05 am
Scott, Rabbi Jonathan Sachs says that based on collective experience of jewish people over last 3,000 years it is safe to assume that all western universalisms regardless of being secular or religious nature are false and anti-semitic. Do you agree with this statement?
Comment #124 May 25th, 2025 at 12:14 am
fun #123: Here’s what GPT4o says about your assertion, and it seems plausible to me—
Rabbi Jonathan Sacks did not assert that all Western universalisms are inherently antisemitic. However, he critically examined how certain universalist ideologies in Western history have, at times, led to the marginalization or persecution of Jews.
Regardless of whatever Rabbi Sacks believed about this, no, I certainly don’t think all Western universalisms are inherently antisemitic. In particular, I see the ideals of the Enlightenment as having led to the emancipation of the Jews, and only various perversions of those ideals (combined with pre-Enlightenment philosophies) as having led to their persecution and murder.
Comment #125 May 25th, 2025 at 2:15 am
Hi Scott, here is Rabbi Sack’s opinion about Enlightenment as per Chatgpt 4o: “The Enlightenment presented European Jews with a messianic promise and a demonic reality. The promise was a secular and rational order in which anti-Jewish prejudice would be overcome and Jewish civil disabilities abolished. The reality was that the more Jews became like everyone else, the more irrational and absolute became the prejudice against them.”
Wondering if you agree with him.
Comment #126 May 25th, 2025 at 4:05 am
fred #118
“ The citizens of America, the supposed greatest “representative democracy” in the world, have no say whatsoever over the fact that their taxpayer money is being used to fund/support Operation Gideon’s Chariots (and this is true regardless of who would have won the last elections).”
I fully support the use of US funds for Israeli operations in Gaza to destroy Hamas stem and root, whatever it takes in the view of the IDF. This was an issue in the last election and so no basis to claim it doesn’t reflect a plurality of US voters. The IDF is exercising my agency in a manner I agree with.
I am not sure what you are suggesting to establish the agency of US voters in Gaza. The IDF have a show and tell each night to discuss tactics for the next day and then direct vote like America’ Got Talent. Please don’t take this as sarcastic just trying to understand what you propose in practice.
Agency and instrumentality to me at its most basic level is the right to provide for self defense. In this case that requires the compete eradication of Hamas. I believe my expectation to be reasonable that if allowed the opportunity they will be back to invading Israel and slaughtering women and children in the most heinous manner first opportunity they have.
This same battle has been in place for three thousand years or so with Arab tribes (and especially Persia strangely enough) the primary aggressors against Jewish tribes, Athenians, Spartans, Peloponnesians, , Christians, Roman Empire, etc. The current events are just a couple of paragraphs in the tome of this history and expect many pages still to be written. It’s just our turn at bat.
The ideas of the Enlightenment etc provide fantastic guidance for well meaning human behavior but were never intended as a national suicide pact in response to violent human actions.
Comment #127 May 25th, 2025 at 6:39 am
OMG #126
I expect that Trump will soon realize that the logical thing to do would be to turn Israel into the 52nd state of the US (right after Canada)… if we’re doomed to be indefinitely forced to be entangled with the region, at least the US would get some of its agency back!
😀
Comment #128 May 25th, 2025 at 11:06 am
fred #127: The idea that Israel is utterly dependent on American support is another canard that inhibits understanding. The assistance that the US provides is a tiny percentage of Israel’s defense budget, and it benefits the US defense industry as much as it benefits Israel (and also, the US gets lots of military and intelligence benefits in return). Even so, I’m increasingly of the view that Israel would be better off without this aid, just because that would cut off one of the central, outsized obsessions of American anti-Zionists.
Crucially, during its defining wars of 1948 and 1967, Israel not only had zero financial support from the US government, but was actually under an American arms embargo (!).
Comment #129 May 25th, 2025 at 12:02 pm
Scott #128
If I and the plurality of US voters can exercise our agency, Israel will never stand alone. This idea about the US controlling Israel overtly had some traction even among Jewish Zionists (more or less) I believe at the start of this. It alarmed me at the time and I am very very pleased it didn’t happen. The US military installation of the dock in Gaza and the provision of supplies was a keystone cops episode. I can’t imagine the bollix now if the US had somehow exercised more control.
Comment #130 May 25th, 2025 at 12:29 pm
Just as an aside, the Merkava is a far superior tank for the modern battlefield than the ballyhooed Abrams. The Abrams has been a total disaster in Ukraine. Australia is sending 80 more of these and the Pentagon is distraught because the effective combat life of the Abrams has been hours to days. It looks bad for $10 million per weapons to be little more than target practice for cheap drones.
The American Maginot Line is high tech horrifically expensive weapons systems. These babies cost $100 million each-they must be effective.
Comment #131 May 25th, 2025 at 1:17 pm
Hi Scott,
I asked for pro-israel score to chatgpt 4o for two triads: Pope-Xi-Aristotle and Lama-Modi-Huemer. It gave a a score of 4 for the first triad and 9 for the second triad out of 10.
Comment #132 May 25th, 2025 at 2:37 pm
Scott -Israel’s defense budget for 2024 was an all time high of 46.5 billion US dollars. From October 2023-October 2024 estimates of military aid provided by the US to Israel was nearly 18 billion. This included a tremendous number of heavy shells and bunker busters. Israel’s capacity to produce these at a scale needed when fighting on multiple fronts is not high enough. Unless those numbers are wrong (they could be-this is just from googling), I’d say that’s far from insignificant.
Comment #133 May 25th, 2025 at 3:24 pm
Scott,
I believe recent events have shown that you were wrong about Trump and Israel before the 2024 election. Do you concur?
I voted for Trump primarily because I thought the Biden administration was shameless in abandoning Israel in their time of need (pressuring them not to go into Rafah, pressuring them not to move against Hezbollah and Iran, basically forcing them to fight with a hand behind their back), and Trump said he was an unequivocal ally of Israel, AND he’d actually do something about the anti-Zionist freaks at the universities, and I believe he’s followed through with both promises.
Read this article in particular (yes, it’s anti-Israel propaganda but still interesting):
https://www.axios.com/2025/05/25/israel-isolated-criticism-gaza-operation-aid
The US is the only Western country still standing with Israel during Operation Gideon’s Chariot. If Kamala (and I won’t even use the insult for her out of respect for this comment section) was president, we likely would’ve pulled the aid by now.
There’s plenty of politicians who pay lip service to Israel, but when push comes to shove, don’t believe that Israel had the right to defend itself and destroy the enemies bent on its destruction, which makes them, in my opinion, antisemites. Because Israel is the only nation held to such an absurd standard.
Comment #134 May 25th, 2025 at 3:40 pm
I’ll leave with a scene from one of my favorite movies, MUNICH
Peace and good luck to everyone.
Comment #135 May 25th, 2025 at 8:27 pm
Scott 123: It’s about 10%, and most of their munitions come from American companies even if they pay for it.
Israel would be in serious trouble without the American alliance, though they could pay for everything they get and be fine(and I think should).
But very few people actually care about the money. American anti Zionists want to cut off the weapons, and the Israelis do really need those.
Comment #136 May 25th, 2025 at 8:38 pm
Fred #134: It is supposed to be the job of the Jews to be righteous; the job of Israel is to secure continued existence of its citizens. However one may aspire for Israel to be the most righteous among nations, wishing for the Jewish state to perish at the semblance of failure in this mission is — on balance — terribly counterproductive methinks.
Comment #137 May 25th, 2025 at 9:20 pm
Henning #101: I agree that Hitler seizing (cf. usurping) power was not necessarily impossible even in the presence of Center-Left unified anti-Nazi coalition. Nevertheless — but this was a secondary point at best by the way — I view Stalin’s order as misguided (Stalin himself is apparently on record as growing to regret it in due course).
Comment #138 May 25th, 2025 at 9:52 pm
Fred #134,
Ehud Olmert disagrees: https://www.yahoo.com/news/indiscriminate-unrestrained-brutal-former-israeli-193347820.html
Comment #139 May 25th, 2025 at 9:58 pm
“The US is the only Western country still standing with Israel during Operation Gideon’s Chariot. If Kamala (and I won’t even use the insult for her out of respect for this comment section) was president, we likely would’ve pulled the aid by now.”
if you think any part of this makes you look good or supports you as morally correct,
you are wrong
Comment #140 May 25th, 2025 at 10:05 pm
Setting all the reasons to be angry aside, I hope we can all care about the over two million people who live in the Gaza strip.
Comment #141 May 26th, 2025 at 3:04 am
Scott #73, Pakistan is founded on the principle that truth doesn’t belong to the second bucket (Judaism-Hinduism-Jainism), while Israel is founded on the principle that truth doesn’t belong to the first bucket (Christianity-Islam-Atheism). If we start with this premise, then motivation and liberality (or lack of it) of many people who support or oppose these two nations become quite clear, imo. What do you think?
Comment #142 May 26th, 2025 at 9:23 am
Concerned #140
Do you actually care about the two million Gazans, or do you just want to stop seeing unpleasant headlines and pictures on CNN? If it’s the former, try giving some thought to Gazans’ likely future if Hamas will remain in power after the war.
Comment #143 May 26th, 2025 at 10:57 am
onlooker #141: No, that’s loony talk. While someone else could say better than me what principles (if any) Pakistan was founded on, Israel was founded on the principle simply that Jews should have a small state in their indigenous homeland that’s committed to safeguarding their lives. Even so, there are many Muslims, Christians, and atheists in Israel (many of the atheists also consider themselves Jews). Just like in the US and other liberal democracies, they’re all free to try to convert observant Jews, as long as they do it by persuasion and not force.
Comment #144 May 26th, 2025 at 11:01 am
Roglic #135: Agreed on both points! Yeah, I meant that Israel would continue buying American munitions but pay full market price for them. This would be a small, easily survivable hit to Israel’s economy, in return for cutting off a huge obsession / talking point of the anti-Zionists.
Comment #145 May 26th, 2025 at 12:03 pm
To all the zioterrorists, zioturds, and the zioterrorist enables and sympathizers (I put Scott Aaronson squarely in this bracket now, thanks for your latest mask-off moment):
I now wholeheartedly support y’all having a safe haven in Israel. Please do us all a favor and move there and cleanse the western societies of your presence. Go and live there happily. Its better to have all the stench concentrated in one place where you all want to be anyway, and keep the rest of the world stench free, so that we don’t become complicit in your immoral murder campaigns.
Also to all the zionists: If “Am Yisrael Chai” is the new “Heil Hitler”. Please scream and cry about it loudly. Cry so loudly so that I can hear you from Antarctica where I’m headed next week. And blame Netanyahu for starting this, since apparently “Free Palestine” triggers y’all so much.
Comment #146 May 26th, 2025 at 1:18 pm
Scott #141, I am trying to define ideologies of two countries by what they are not rather than what they are. It actually clarifies the matters, imo. And there is not much disagreement between you and me except the point about Atheism, imo. While ideology of Pakistan was reiterated by its Army chief just one month back, and while Israel is certainly not an Islamic or a Christian country, and for good reasons, it is even more against atheism, and again for a good reason, imo. While chatgpt is your friend to confirm what I said, I will just mention one fact in support of my argument: Civil marriages are not allowed in Israel.
Comment #147 May 26th, 2025 at 1:26 pm
onlooker #146:
Civil marriages are not allowed in Israel.
They’re not performed in Israel (something that I join many liberal Israelis in wanting to see changed), but they’re recognized there. In practice, that means that anyone wanting a civil marriage just flies to Cyprus or some other country for the marriage ceremony.
Comment #148 May 26th, 2025 at 1:34 pm
MIZH #145:
To all the zioterrorists, zioturds, and the zioterrorist enables and sympathizers (I put Scott Aaronson squarely in this bracket now, thanks for your latest mask-off moment):
I now wholeheartedly support y’all having a safe haven in Israel. Please do us all a favor and move there and cleanse the western societies of your presence.
MIZH, even while I leave a huge amount of your openly antisemitic garbage in moderation, I’m letting the above comment through in order to make the following point: this is why Israel needs to exist. Because, however well-integrated Jews become in countries like the US, however much they contribute to those countries, there will always be a large fraction of their fellow citizens who think as you do. And history has abundantly shown that we need a place to escape you and your friends every now and again.
Comment #149 May 26th, 2025 at 2:45 pm
MIZH #145, the ratio of deaths linked to state actions between Pakistan and Israel is approximately 35:1. So, why Pakistanis and its sympathizers get a pass while Zionists and its sympathizers are criticized endlessly?
Comment #150 May 27th, 2025 at 8:19 am
Israel is illegal state i.e. in other word, illegitimate state, and I can prove it.
If we look carefully on the rules of international law, we can see that the creation of the State of Israel, and also many of things what happen after, are not matching with legal standards. Some people are saying, if a state exists and is recognized by other states, then it must be legal. But this is not correct. Law is not only about who exists, but also how this state come into existence, and how it is behaving now. And in the case of Israel, both the origin and the present are showing serious problems.
One of most important principles is right of self-determination. It is written in Article 1 of United Nations Charter, and also in international treaties like the Covenant on Civil and Political Rights. This principle say that all peoples have right to decide for themselves about their future and government. In 1947, the UN General Assembly made Resolution 181, which recommended that Palestine should be divided into one part for Jews and one part for Arabs. But this was only recommendation, not something what had legal force. And more important: majority of people in Palestine at that time were Arabs, around 65%. They did not agree for this partition, so it was against self-determination. History or religion cannot replace legal right of people who live in the land now.
After Israel declared its independence in May 1948, there was war. During that war, more than 750,000 Palestinians were forced to leave or escaped from their homes. This is what Palestinians call the Nakba. Many people think this was just result of war, but Israeli archives show that it was done with planning, to create Jewish majority. This is serious problem. Article 49 of the Fourth Geneva Convention says it is forbidden to forcibly move population from the land. Also, the United Nations made Resolution 194, which says that Palestinian refugees have the right to return. Israel never allowed them to come back, and this violation continues until today.
Not only past is the issue. Also now, Israel is violating many rules. Since 1967, it is occupying the West Bank and East Jerusalem, and continues to build settlements there. This is not allowed. UN Security Council Resolution 242 says that it is not legal to take land by war. The International Court of Justice also said in 2004 that the occupation is illegal, and that the wall and settlements are breaking the law. According to Rome Statute, transferring own population into occupied land is war crime. Still, Israel is doing this and no one is stopping them.
Another very serious issue is apartheid. This word is not just political insult, it is legal term, and it is crime under international law. It means a system where one group rules over another group in systematic way, using laws and institutions. In Israel and in the occupied territories, there are different laws and rights depending if person is Jewish or Palestinian. Palestinians have less freedom, worse access to resources, and often no legal protection. This is not just opinion—it is conclusion of Amnesty International, Human Rights Watch, and United Nations experts. Apartheid was illegal in South Africa and it is illegal now too. A state cannot be legal if it is doing apartheid.
Sometimes people say Israel is legal because it has population, territory, government and can do foreign policy. These are the Montevideo criteria. But this is only minimum for practical statehood. It doesn’t make everything legal. There is rule in law: ex injuria jus non oritur—from illegal act, no legal right can be born. If a state is born from expulsion, occupation, apartheid and breaking self-determination, then we cannot say it is fully legal, even if other countries recognize it.
Saying Israel is illegal is not about hate or denying Jewish people right to live. It is about law, and applying it same way to all. If international law means something, it must apply also to powerful countries. If we ignore it just because it is difficult, then we destroy the idea of justice and rules. We make law something what works only for weak countries, not for all.
Comment #151 May 27th, 2025 at 1:58 pm
Aleksy #150: Thank you for finally resolving a longstanding open question, via your proof of Israel’s illegality! I’m sure that all 10 million Israelis, after they’ve read and understood their proof, will respond in the only possible way—by committing mass suicide, like the Jews of Masada 2000 years ago.
Too flippant? No, that’s literally what’s at issue here. When you say that the majority of residents of Palestine in 1948 rejected the UN’s partition proposal, you somehow neglect to mention that their alternative proposal was literally to murder every last Jew there. They failed, but not for lack of trying. You also somehow neglect to mention that, for every Arab who fled or was expelled from Mandate Palestine, a Jew was expelled from the Arab world. Indeed, Israel allowed Arabs to stay who wanted to live peacefully in the new state of Israel (they’re the descendants of the now 2+ million Israeli Arabs), whereas most Arab nations extended no similar courtesy to their Jews.
The other thing you neglect to mention is that your email address is from Poland—the country where, a few years before Israel’s founding, millions of Jews, their very existence having been declared illegal, were sent to the gas chambers, with the support and cooperation of most of the Polish population.
In that case, if you were a Jew, what would you do: acquiesce when someone said the law required your lungs to be filled with Zyklon B? Or say “fuck your interpretation of the law then,” and fight for your life?
If the latter: how is the situation any different when considering whether modern Israel gets to exist or not?
(And finally: which countries on earth could survive the standards you’ve invented for Israel here? Which ones were founded fully in accordance with laws that existed at the time?)
Comment #152 May 27th, 2025 at 2:19 pm
Scott, Aleksy is trolling you. It’s just barely possible that he actually believes all this “international law” drivel himself, but there’s no way he thinks there’s a reasonable chance of you finding this persuasive; ergo, troll.
Comment #153 May 27th, 2025 at 2:23 pm
Scott,
Thank you for your reply. I see clearly you feel strong emotion on this topic. That is understandable. But emotion is not same as argument, and I will respond to your points with facts and logic. I do not write to provoke, but also I will not apologize for saying what I believe is true and supported by evidence.
You begin with sarcasm—about all Israelis reading my comment and committing suicide like at Masada. I don’t think this helps serious conversation. I never said Israel should not exist, and I never called for violence against anyone. I made argument from legal perspective, based on international law. Israel exists. Yes. But existing is not same as being legal. Many regimes existed while violating law: apartheid South Africa, colonial Algeria, Nazi Germany. So your sarcastic reaction avoids the question.
You claim that the Arab population in 1948 rejected the UN Partition Plan because they wanted to “murder every last Jew.” This is simply not accurate. Yes, there were tensions, violence, and yes, extremist rhetoric. But the main reason Arabs rejected the plan is because it gave 55% of the land to the Jewish state, when Jews were only about 30% of the population and owned less than 7% of the land. Any people in the world would reject such deal. That is not “genocide,” that is saying no to foreign-imposed partition.
The Zionist movement was not just peaceful either. The Irgun and Lehi militias carried out terrorist attacks. The Deir Yassin massacre happened before Arab armies invaded. Plan Dalet, a strategy for conquest and depopulation of Arab villages, was already written. And 750,000 Palestinians were expelled or fled—often by force or threat—not because they wanted to kill Jews, but because their homes were under attack. The ethnic cleansing is not imaginary; it is documented in Israeli historian Benny Morris’s own work.
You say that for every Arab expelled from Palestine, a Jew was expelled from Arab lands. This is again misleading. First: the expulsions of Jews from Arab countries were tragic and unjust—but they were not caused by Palestinians. Second: many of those Jews were welcomed by Israel because it needed population to replace the Arabs it had expelled. This does not make it morally even. Two wrongs don’t make a right. If Arab states mistreated Jews, they too should be condemned. But this cannot be used as justification for ethnic cleansing of Palestinians.
You also say Israel allowed Arabs to stay, and now 2 million live in Israel. That is true. But they live under more than 65 laws that discriminate against them. They were under military rule inside Israel until 1966. They are not allowed to return to their villages destroyed in 1948. And in the occupied West Bank and Gaza, millions live under full military control. This is not equality. This is apartheid—defined not by me, but by UN experts, by Human Rights Watch, by Amnesty International.
Then you bring up Poland. You write that I use Polish email address, and you accuse Poland of helping kill Jews in gas chambers. That is a very serious and false accusation. Poland was invaded and occupied by Nazi Germany. Six million Polish citizens died in the war—half of them Jews. Yes, some Poles collaborated with Nazis. That is shameful. But many others resisted. Thousands of Poles hid Jews. Yad Vashem recognizes more Righteous Among the Nations from Poland than from any other country—over 7,200. Poland had no collaborationist government like Vichy France or Slovakia. Auschwitz was not built by Poles. It was built by Germans on occupied Polish soil. I do not deny Polish antisemitism—but don’t rewrite history to blame the victims of occupation for the crimes of the occupiers.
And finally, you ask what I would do if I were a Jew told by law to go to the gas chamber. I say very clearly: I would resist. But this has nothing to do with Palestinians asking for rights under international law. Israel is not resisting genocide—it is imposing occupation. No one is trying to gas the Jews in Tel Aviv. But in Gaza, whole neighborhoods are bombed, children are starved, electricity and water are cut. This is not self-defense. This is punishment of entire people.
You ask which states were founded in perfect legality. Fine—many states have blood in their history. That does not mean we give up on legality. If we say no state must ever follow law, then why do we have international law at all? Why have Geneva Conventions? Why punish war crimes? Why have UN resolutions? We should strive for higher standard, not abandon it because it is inconvenient for some.
I’m not asking you to agree with me. But if you want to answer arguments, do it with facts, not sarcasm, not slander, not Holocaust guilt-trips. I respect your intelligence. I expect you to use it.
Regards,
Aleksy
Comment #154 May 27th, 2025 at 3:23 pm
Aleksy #150:
There is a lot of anti-Israel propaganda floating around. Some of it is outright lies; some of it intentional mischaracterizations. Anti-zionists tend to dismiss any facts that contradict their views, but perhaps you would be willing to take a look at the statistics posted by Hamas? The data most favorable to Hamas, cherry-picked, sanitized, and published by Hamas for the sole purpose to “prove” to the World that Israel is committed genocide? Here it is:
32,152 children in Gaza lost their fathers.
4,417 children in Gaza lost their mothers.
Now, ask yourself: is this statistics, even if taken at face value, consistent with genocide, with deliberate killing of civilians, women and men alike? Why does the number of men of fighting age (fathers) who perished in this war outnumber the women 7.2:1?
Genocide, deliberate murder of civilians, is what happened on October 7. What’s happening today in Gaza is war, with minimal number of civilian casualties, as far as wars go: in WWII 60-67% casualties were civilians; according to the Hamas orphan data it’s 12-15% in Gaza. Initially Hamas claimed that civilian casualties were 70-80%, but slip-ups like this one in their own reports disprove that.
https://x.com/QudsNen/status/1882406281278796284
Comment #155 May 27th, 2025 at 3:48 pm
Vladimir,
I must say your comment is disappointing. Instead of answering even one point I raised, you decide to just call me a troll. Is this really the best you can do?
I wrote long argument, with references to international law, to UN resolutions, to historical facts. You can disagree, fine—but then you should explain why I am wrong. Calling it “drivel” is not argument. It is avoidance.
If you think international law is meaningless, then say it directly. But then be honest—don’t pretend to care about legal norms in any other context either, whether it’s Ukraine, or Syria, or war crimes somewhere else. If law only matters when it is convenient, then it is not law anymore—it is propaganda.
And to say I am a troll just because I don’t expect Scott to agree? That is strange logic. We write things not always because we think someone will be converted, but because ideas must be challenged. If you think his position is so strong, then it should survive criticism. If it cannot, maybe it needs to be re-examined.
So if you have something real to say—about self-determination, or ethnic cleansing, or legality of occupation—I’m listening. If not, then calling names is not serious discussion.
Comment #156 May 27th, 2025 at 4:00 pm
Michael P,
Thank you for your message. I read it carefully, but I must say—I really don’t follow the logic here.
You accuse others of spreading propaganda, of cherry-picking, and then you give one piece of data and say it proves that what happens in Gaza is not genocide. I don’t think it works like this. You write that 32,152 children lost fathers, and 4,417 lost mothers. Then you ask—how can this be genocide, if more men died? But first of all: who said that genocide must kill men and women in exactly same proportion? This is not how genocide works. In many genocides, like Bosnia or Armenia, men are killed first, exactly because they are seen as potential fighters or protectors of community. This doesn’t make it less genocide. It makes it more targeted.
Second, you assume that father means fighter. But that is not serious claim. Most men in Gaza are civilians. They are teachers, farmers, drivers, store owners. And in Gaza, it is mostly men—fathers—who leave house to search for food, for water, for fuel, or to help neighbors.They are the ones who leave house for work and making money. They go outside during bombardments because they feel responsible. Of course more of them die. This does not make them military targets. It makes them victims of impossible situation.
And if over 30,000 children lost fathers, and over 4,000 lost mothers, this shows clearly that civilian society is being destroyed. These are families—not combatants. These are homes, not bunkers. What kind of future these children have now? You think this is normal in “just war”?
Then you say that in WWII, the percent of civilian deaths was higher, so Gaza war is “minimal.” But in WWII, most civilians could escape. Trains worked. Borders were open. Aid could come. In Gaza, nobody can leave. Borders are closed. Hospitals are bombed. People are drinking salty water. So yes, percentage may look different—but the condition is worse. You also say October 7 was genocide. No. It was massacre. It was horrible, yes. But it was not systematic campaign to erase whole population. What happens now in Gaza is systematic: homes flattened, universities erased, journalists killed, food supply destroyed. This is not defense. This is collective punishment. Which is illegal under international law.
You want to use numbers? Then look at all numbers. Not only one post that looks convenient. Look how many children are dead. How many schools are gone. How many people eat grass and drink dirty water. This is not “war with low civilian cost.” This is something different. Something we should name clearly.
If you disagree, okay. But don’t pretend that killing 30,000 fathers is evidence of restraint. It is evidence of something else.
Comment #157 May 27th, 2025 at 4:37 pm
If anyone still follows Peter Woit’s blog, then his last comments seems deranged moving into the realm of conspiracy theory, where it is not the protesters that handed out Hamas’ materials, but some agents provocateurs. It seems that even other people from Columbia are in disagreement with his attempts to deny any wrong doing by Columbia. I wonder how much his views represent other staff members. If they do, maybe Columbia should be closed down.
Comment #158 May 27th, 2025 at 4:41 pm
Alexy #155,
How about killing 8,000,000 Germans? Were Allies at fault for that? Or maybe Nazis who started tge war? Speaking of that war, for an Eastern Europen to claim that civilians could escape, on either side, is disingeneous. Like 99.7% Jews of Poland.
There are several wars going on as we speak, several more bloody than Gaza. And yet leftists are focussing on Israel, a country that didn’t start the war, willing to end the war when hostages are returned and murderers surrender to international law, and doing utmost to minimize civilian casualties.
I don’t believe that you don’t understand this. You came with an agenda, just like most of the left, and logic is futile here.
Comment #159 May 27th, 2025 at 7:10 pm
Yiftach #157
> If anyone still follows Peter Woit’s blog, then his last comments seems deranged moving into the realm of conspiracy theory, where it is not the protesters that handed out Hamas’ materials, but some agents provocateurs
All according to schedule:
https://scottaaronson.blog/?p=8853#comment-2009829
Comment #160 May 28th, 2025 at 2:52 am
Aleksey#141
It would take considerable time to answer you in detail but your entire argument is in practice a house of cards although it may be consistent in theory. Just as one example-
“ It is written in Article 1 of United Nations Charter, and also in international treaties like the Covenant on Civil and Political Rights. This principle say that all peoples have right to decide for themselves about their future and government”
In order to form the United Nations Stalin was effectively ceded control of Eastern Europe (including Poland of course) as a buffer zone. Were the peoples of Eastern Europe allowed to determine their future and their government as part of the formation of the United Nations itself? A paradox on delivery.
The belief that sprinkling the world with Western liberal democracy would eliminate international conflict is flawed. It hasn’t worked and the definitive experiment to test this was sprinkling a trillion dollars of these beliefs on Iraq by the Bush Administration.
Even accepting everything you write there is a folk wisdom adage that applies when law comes into conflict with self defense at the level of an individual-I would rather be judged by 12 (size of US jury) then carried by 6 (pall bearers).
As an aside Cordell Hull received the Nobel Prize for establishing the UN. Hull was an anti-Semite that would be burning in Hell right now for refusing the MS St Louis entry to the US if in fact there is a Hell in accordance with Christian sin theory but in fact he won the Prize.
Comment #161 May 28th, 2025 at 4:27 am
I hate using this analogy because it has been so mis-used but these distinctions between ant-Semites and anti-Zionists on college campuses do not make sense to me.
I don’t see the distinction between the Nazis and Hamas. To me Hamas are the same anti-Semites just without the instrumentality of the Germans. People that supported the Nazi Party in Germany are in practice called Nazis and anti-Semites. On college campuses they are called anti-Zionists. In my view this is mostly just the usual language game of the Left. I agree with posters above like Yiftach and Julian that pro-Hamas is certainly anti-Semitic and this also applies to anti-Zionists in the general case (I believe I am missing another poster that made this point and so my apologies).
As Dr Aaronson points out there must be a Jewish state and opposing a Jewish state in general implies antiSemitism. You could make the argument that the world is flawed in that Jewish people NEED a Jewish state and I would agree. The world is flawed but is what it is.
The Left’s argument seems to be based on-forget the three thousand years of history, and have no regard for the future, but look at the conditions today alone. Good advice if you have no memory and are unable to develop reasonable expectations.
Comment #162 May 28th, 2025 at 8:10 am
Congratulations Scott for this great achievement!
Regarding the usual Israeli-Palestinian war argument:
Some historians believe that the reason for WW2 was that Germans were able to deny their defeat at WW1. Anything happening in Gaza today is to prevent the next war which would be even bloodier. The ones that brought all the violance are the people who funded Palestinian Schools for the past 30 years, teaching them to hate Jews and that violent would lead them to victory( EU US and UN). These silent self rightous murderers are the ones that should be ashamed of the results of the seeds of violence they nurished for decades.
Comment #163 May 28th, 2025 at 10:08 am
Scott,
Just small note—I noticed you didn’t reply yet to my previous message. I understand of course if you are busy or not interested to continue the discussion. Still, I think some of the points I raised deserve proper answer, especially since your first reply was very strong and personal.
If you decide not to respond, okay. But then let it be clear: it is not because the arguments were unserious. It is because they were difficult.
Regards,
Aleksy
Comment #164 May 28th, 2025 at 11:11 am
There was a recent debate between Aiden Doyle and a Palentinian terrorist, where Aidan obliterates the terrorist’s talking points. It is very apt to add this to Aleksy’s arguments.
The longer video is interesting to learn more about Israeli point of view, and contains testimony of actual hostages.
Highlights:
https://www.youtube.com/watch?v=p_gz7gIaCNc&ab_channel=TheDailyReminder
Full version:
https://www.youtube.com/watch?v=jmkh39WkZBc&ab_channel=TheGr8Debate
Comment #165 May 28th, 2025 at 12:07 pm
Yiftach #161
I don’t see how people would be expected to be anti-Hamas if they were only allowed to see Hamas-linked materials that had been filtered through our western side, which they would call biased in a way that was impossible to effectively dispute. If I wanted to rally support for Israel’s existence, I’d go and find an indisputable agent of Hamas and have them hand out their own materials.
You will notice that Likud, being a lot smarter and more effective than Hamas as a supporter of violent response to the fact that there is more than one religion Earth and more than one claimant in most lands, has not shown US students any of their materials, or made any effort to translate their vision to English.
Comment #166 May 28th, 2025 at 3:43 pm
Aleksy #163: Please don’t take this the wrong way, but—I feel like the world would be a better place if you were not part of it.
My reasoning is as follows: when you single out a single UN member state, among all ~200 of them, as being illegitimate and having no right to exist—when, moreover, that state is literally the only thing standing between half the world’s Jews and their violent deaths in a second Holocaust—when it’s obvious to any fairminded person that, if you applied 10% of the same level of legal scrutiny to the founding events of other countries, the UN General Assembly would need to be swept nearly bare—when, finally, you needle a productive scientist over and over, commenting and emailing to ask why he hasn’t replied to you yet, taunting him that if he doesn’t then he’s effectively conceded the argument, etc. etc.—it’s clear that you’re making a negative contribution to the world.
I feel like, if you understood this, you’d see that the right and honorable thing to do would be to kill yourself as quickly as possible.
I hasten to add, however, that I’m not saying this out of any personal animus whatsoever towards you. It’s purely disinterested reason that’s led me to these conclusions. If you respond to me emotionally rather than rationally, that will show me that (alas) you weren’t ready for a logical, evidence-based discussion of these matters.
Comment #167 May 28th, 2025 at 7:01 pm
Scott 166: You should reread your comment. I don’t disagree with the idea at all, but suggesting suicide makes you look bad regardless of what came before.
Just ignore those people. Lots of people have genuine crazy beliefs about Israel, but I’m pretty confident they aren’t reading your blog and commenting. He’s a troll.
Comment #168 May 28th, 2025 at 9:42 pm
Roglic #167: I mean, Aleksy had called for my family members to die. That is what it means, in concrete practical terms, for Israel’s existence to be criminal and illegitimate: that there should no longer be anything standing between 7 million Jews and forces whose entire political program consists of the murder of those Jews.
That’s pretty bad already, but then, incredibly, Aleksy made it even worse by gaslighting me, pretending that he’d reached his stance by pure rational deliberation without any hint of animus, and that I was honor-bound to discourse on the matter according to his rules for as long as he liked. Which immediately raised the question: how would he react if someone called for his death in the same pompously faux-rational tone?
So, while it takes a lot for me to respond as I did, Aleksy cleared the bar. I want him to know that I see him as no different, not even slightly, from his many countrymen who eagerly helped the Nazis send my relatives to the gas chambers.
Comment #169 May 29th, 2025 at 11:12 am
When I read Woit’s accusations I am reminded of a principle in Alinsky’s handbook For Radicals-Accuse them of what you are doing and deny deny deny. I won’t further characterize his accusations and will leave this simply as a statement of fact.
Comment #170 May 29th, 2025 at 1:13 pm
OMG #169
> Accuse them of what you are doing and deny deny deny
Yup. E.g. in the aftermath of the Columbia library takeover three weeks ago, Woit claimed that Scott would have been “welcome to come study and learn the error of your ways just like anyone else” in the library, and when confronted with video evidence to the contrary, denied hearing the words “don’t let this guy in, he’s a fucking Zionist” in the video, despite being provided with the exact time they’re shouted.
Comment #171 May 29th, 2025 at 9:38 pm
I simply don’t see an endgame in Gaza that doesn’t involve Israeli re-occupation. Defeating Hamas and withdrawing will only allow another violent antisemitic group to take its place. The occupation of Gaza is necessary for Israeli security. And of course ending the West Bank occupation is simply beyond the pale—the next October 7 planned from the West Bank could ve a thousand times worse.
As much as we might hope for a two-state solution, practical realities make it impossible, for the next couple decades at least. Any withdrawal of Israeli forces from occupied territories poses a security risk. The only possibility at this moment is a “one-state solution”—Israeli occupation of West Bank and Gaza.
Comment #172 May 30th, 2025 at 6:42 am
Vladimir #170
I could certainly hear it. Strange sample of students in your video. I checked and there are 200 more female undergrads at Columbia than males but the video shows nearly all females and the voice telling him to *get the f out of here” is a female voice. I guess the anti-Zionist and anti-patriarchal belief sets have merged before our eyes.
Julian #171
I agree with your point but 3000 years of history suggests “next couple of decades” likely to be off by a couple orders of magnitude. Humans have the capacity to form belief sets that are impervious to empiricism. (I can think of no bigger explicit rejection of empiricism than the belief that no difference between athletic performance of males and females). At that point reason is not a viable tactic to resolve conflict. They are certainly welcome to have any belief set they choose but when it results in hostile existential action against others then you accept their beliefs or fight for survival and a reasonable expectation of future safety. In this case I don’t understand the objection to resettlement as an option. In the long run it would save not only Israeli lives but Gazans as well. It seems to me that a buffer zone of 100’s of kilometers far better than meters.
Comment #173 May 30th, 2025 at 9:37 am
Scott,
Did you see this shameful display at MIT commencement? The student president, Megha Vemuri, decided to spend most of her commencement speech moralizing about Palestine:
https://x.com/KassyAkiva/status/1928233466627318135
Sorry for the bad news, but it’s important to stay informed. MIT has made some big pro-merit actions recently (cynically, to appease the admin … optimistically, because they are following the zeitgeist)… shutting down their DEI office and making a website (understanding.mit.edu) showing how cool their research is and highlighting that they are “merit-based”. I feel like this could be a setback to that, though maybe it will pass unremarked….
Comment #174 May 30th, 2025 at 10:06 am
There’s a lot of hype out there about “quantum algorithms for machine learning,” but as far as I understand, in most cases, these only offer quadratic Grover-type speedups, or it’s dubious that there’s not a classical heuristic matching their performance.
I have a question about something else in this domain, though. Namely, if P=NP, and there’s a fast algorithm for an NP-complete problem, what would this imply in ML? Would it offer dramatic speedups in training ML models e.g. by gradient descent, or even faster than gradient descent, by finding global minima in polynomial time? Is training ML models an NP optimization problem? My intuition says it is. But I’m not familiar with CS theory at all, so go easy on me!
Comment #175 May 30th, 2025 at 10:42 am
Vladimir #170
It seems that Dr. Woit is acting consistently contrarian. With string theory, Woit maintains the contrarian viewpoint, arguing against the “powers that be” and positing that their theory is inadequate. Woit seems to have extended this stance to middle eastern politics, opposing the supposed “powers that be” at Columbia that have “aligned with Israel.” So perhaps in Woit’s view, this contrarian viewpoint is just a natural extension of his rebellious nature, and this is why he sees nothing wrong with it. However, it is disturbing that Woit does not seem to grasp the similarity of his recent rhetoric with nefarious organizations (past and present). Moreover, given Woit’s reputation for criticizing string theorists for making claims with lack of evidence, it is ironic that Woit seems to lack evidence for his recent claims both about Columbia and about Scott (e.g. using various straw man arguments against Scott, and mis-characterizing Scott’s positions).
Comment #176 May 30th, 2025 at 1:10 pm
@Julian #174:
If the scoring function can be computed in polynomial time, then finding the optimum is in NP. This wouldn’t follow if the scoring function is given by an oracle, though.
Comment #177 May 31st, 2025 at 9:44 am
Some of your viewers might be interested in the quantum computing article in this year’s SIGBOVIK. It runs Shor’s algorithm on a noisy hardware quantum computer and factors small numbers.
> Historically, most papers that claimed they “ran Shor’s algorithm” didn’t run Shor’s algorithm. It’s unfortunately common to run circuits *inspired by* Shor’s algorithm, but with key pieces replaced by trivial pieces. The issue is that, in the large shenanigans limit, all quantum factoring circuits are trivial. In this paper, I don’t make that mistake. I run Shor’s algorithm exactly as it was supposed to be run: with comically underoptimized circuits. Nevertheless, the circuits works. They quickly result in factors being produced. This naturally raises the question: “What’s the catch?”. Because of course there’s a catch.
> […]
> To my knowledge, no one has cheated at factoring in this way before. Given the shenanigans pulled by past factoring experiments, that’s remarkable.
Craig Gidney, “Falling with Style: Factoring up to 255 ‘with’ a Quantum Computer”, (2025-01) in ed. the Association for Computational Heresy, a record of the proceedings of SIGBOVIK 2025 the nineteenth annual intercalary robot dance party in celebration of workshop on symposium about 2⁶th birthdays; in particular, that of Harry Q. Bovik, “http://sigbovik.org/2025/” .
Comment #178 June 1st, 2025 at 9:26 am
Joe #175
Ironic yes but I don’t believe consistently contrarian. His current statements are unfortunately largely consistent with the super majority beliefs certainly in the Ivies and also in US academia in general.
Comment #179 June 2nd, 2025 at 12:17 pm
Scott, there is no liberal case for Israel. And it doesn’t prove failure of Israel but proves the failure of liberalism, imo.
Comment #180 June 6th, 2025 at 12:25 pm
As someone else who receives such emails about how they’ve figured out the secrets of the quantum universe, I have *also* noticed the uptick in the inclusion of ChatGPT transcripts. So bizarre!
Comment #181 September 25th, 2025 at 10:12 pm
[…] business problems” detached from any underlying truth-claim. And even here at one of the top 50 quantum computing blogs on the planet, there’s nothing I can do about it other than scream into the […]