I assume you're using the "regular" Pro version of Gemini 3.1 for the above, rather than the Deep Think mode, which is more comparable to GPT-5.5 Pro. To my knowledge, regular 3.1 Pro is a tier below and often makes mistakes.

Moreover, there's no reason to believe the progress of LLMs, which couldn't reliably solve high-school math problems just 3–4 years ago, will stop anytime soon.

You might want to track the progress of these models on the CritPt benchmark, which is built on *unpublished, research-level* physics problems:

https://critpt.com/

Frontier models are still nowhere near solving it, but progress has been rapid.

* o3 (high) <1.5 years ago was at 1.4%

* GPT 5.4 (xhigh), 23.4%

* GPT-5.5 (xhigh), 27.1%

* GPT-5.5 Pro (xhigh) 30.6%.

https://artificialanalysis.ai/evaluations/critpt.

I'm no physics professor but this aligns with the way I use the tools in my "senior engineer" space. I bring the fundamentals to sanity-check the trigger-happy agent and try to imbue other humans with those fundamentals so they can move towards doing the same. It feels like the only way this whole thing will work (besides eventually moving to local models that do less but companies can afford).

LLMs are at their best when you have an expectation for their output. I generally know the shape of the correct response and that allows me to evaluate it's output on it's "vibes", rather than line by line. If there's no expectation then I have to take everything at face value and now I'm at the mercy of the machine.

please, sign up for a paid plan of either chatgpt or claude. gemini is while close, still noticeably behind

you deserve opinions shaped by interactions with the best tools that are out there.

This is close to my experience with code. LLMs can pick out small mistakes from giant code changes with surprising accuracy, or slowly narrow down a weird. On the other hand I've seen them bravely shoulder on under completely incorrect conceptual models of what they're working with and churn around in circles consequently, spin up giant piles of slop to re-implement something they decided was necessary, but didn't bother to search for, or outright dismiss important error signals as just 'transient failures'. Unlimited stamina, low wisdom.

Chiming in to agree but clarify that the latest sota models are no better than Gemini.

I put my stuff through several sota models and round robin them in adversarial collaboration and they are all useful even though, fundamentally, they don’t “understand” anything. But they are super useful delegates as long as deciding on the problem and approach and solution all sits safely in your head so you can challenge them and steer them.

So I know the article is about one particular new model acing something and each vendor wants these stories to position their model as now good enough to replace humans and all other models, but working somewhere where I am lucky enough to be able to use all the sota models all the time, I can say that all keep making obvious mistakes and using all adversarially is way better than trusting just one.

I look forward to the day one a small open model that we can run ourselves outperforms the sum of all today’s models. That’s when enough is enough and we can let things plateau.

I've been watching the automation of things like flight control systems for the past decade, and the evolution of the fallback to a real pilot in the event of a emergency is what's most concerning about where LLMs are being embedded.

Right now, we have a lot of smart people who have trained for decades to understand where these things go wrong and how to nudge them back, but the pool of people are going to slowly be replaced by less knowledgeable.

At some point, a rubicon will be crossed where these systems can't fallback to a human operator and will fail spectacularly.

> by solving hard problems you get an insight into the problem-solving process itself, at least in your area of expertise, in a way that you simply don’t if all you do is read other people’s solutions. One consequence of this is that people who have themselves solved difficult problems are likely to be significantly better at using solving problems with the help of AI, just as very good coders are better at vibe coding than not such good coders

Yes but it's not just that if you solved a problem yourself, you're better at solving other problems; it's also that you actually understand the problem that you solved, much better than if you simply read a proof made by somebody (or something) else.

I see this happening in the enterprise. People delegate work to some LLM; work isn't always bad, sometimes it's even acceptable. But it's not their work, and as a result, the author doesn't know or understand it better than anyone else! They don't own it, they can't explain it. They literally have no value whatsoever; they're a passthrough; they're invisible.

I feel like you slightly miss both points.

> Training must start from the basics though.

Sure, but the point is that at some point (e.g. when starting a PhD) one needs to do research, not learn the basics. And LLMs make that harder, because they solve the "easy research" part.

Take a young lion "fighting/playing" with another young lion as a way to learn how to fight, and later hunt. And suddenly they get TikTok and are not interested in playing anymore. Their first encounter with hunting will be a lot harder, won't it?

> People pay coders to build stuff that they will use to make money and I can happily use an AI to deliver faster and keep being hired.

Again, that's true but missing the point: if you never get to be a "good coder", you will always be a "bad vibe coder". Maybe you can make money out of it, but the point was about becoming good.

But perhaps we should regard it as a major achievement.

@NotOscarWilde drop your email here, I will reach out and happy to get you a pro account for a few months so you can try 5.5 pro.(work at OAI)

At my university, everyone had to pay their AI subscriptions out of their own pocket, until a communal AI service was introduced recently. It took 2 years to set up and only serves gpt-oss-120b, so everyone is still using other services. But at least some admin can scatter the word "AI" all over the university's website now and has an excuse to reject any requests for AI subscriptions because "we already have AI".

It’s a classic example of the best positioned people being in the best position to keep reaping all the rewards.

There’s the example of a poor person and a rich person buying boots. The poor person’s boots wear out and have to be replaced while the rich persons boots last for many years due to higher quality craftsmanship. Over years, the poor person’s boots wear will pay may for boots.

OpenRouter lets you pay by the token only (no subscription), has all the frontier models (including Opus 4.7, GPT-5.5) and most of the others, and if you use it sparingly it usually turns out to be quite cheap.

I fully understand your rant! I pay ~20€/month for the Pro account, as my university has a deal with Microsoft and only seems to recognize Copilot, so it’s very hard to use one own’s funding for paying something else.

Paste what you want me to ask 5.5 Pro and I'll paste you the response.

I replied to a comment about AI in sports and I build on that.

We praise car drivers despite most of the performance in their sport comes from the car. The driver makes the difference when two cars are close in performance. Brilliances or mistakes. Horse riders too.

In the case of math, the human can lead the LLM on the right track, point it to a problem or to another one. So it deserves some praise.

Then the team that built the car, cared about the horse, built the AI might deserve even more praise but we tend to care more about the single most visible human.

It may not be a major achievement by the mathematician (although it's debatable) but it would still be a major result.

I don't know that it's that disappointing. I doubt most of the great mathematicians were actually doing it to achieve immortality. I suspect most of them were either after (possibly indirect) practical applications (via the math -> physics -> engineering pipeline) or just "for the love of the game", appreciation of the beauty of math and the intellectual joy of doing it. AI might also take over the practical application side, but the other aspects are still there for the taking.

Now repeat that for every sort of human achievement

As someone who is much further down the track, I would kindly suggest you drop that line of thought. I've seen far too many brilliant and ambitious people drop into depression because of it.

You are worthy of doing this work because you are able to do it. Do the work because you love it and because you love the mystery. Enjoy every moment that you get to do it. Find joy in the great fortune you have to do this work while others toil away on tasks that bring them no satisfaction. Sometimes it's tedious, but sometimes it's incredibly rewarding in its own right.

Don't work for the possibility of eternal glory though, it just doesn't exist anymore.

You are worthy. You will hone your skills in grad school and be able to command these AIs better than somebody who hasn’t struggled with hard problems for a long time.

I feel my input to LLMs is most valuable in the initial idea, big picture design tweaks, and the vast majority of my usefulness is negative feedback. This looks wrong, you've gotten off track, you're cheating with workarounds, you're falling into a rabbithole, etc.

That's the top link (i.e. that the title is linked to), no?

> but we need to avoid putting their works in production, or in front of other humans, without assessing it by any possible mean.

I think this is good advice in general, maybe with an emphasis on public vs. private, friendly contact. Having 0 thought AI slop thrown at you out of the blue is rude. "could have been a prompt" indeed. But having a friend/colleague ask for a quick glance at something they know you handle well is another story for me.

If I've worked on a subject for a few years, and know the particulars in and out, I'd have no trouble skimming something that a friend or a colleague sent me. I am sparing those 5-10 minutes for the friend, not for what they sent. And for an expert in a particular domain, often 5 minutes is all it takes for a "lgtm" or "lol no".

> "After 16 minutes and 41 seconds, it came back" ... "further 47 minutes and 39 seconds" ... "After 13 minutes and 33 seconds" ... "After 9 minutes and 12 seconds" ... "After 31 minutes and 40 seconds" ... plus other computations Anyone spotting the issue here? What did that really cost?

Whatever the Joules... (convert to $ using your preferred benchmark price) it is a fraction to what it might take a human Ph. D. weeks to feed and sustain themselves when working on the same problem. The economics on LLMs is just unbeatable (sadly) when compared to us humans.

Still not as bad for the environment as animal agriculture, and animal agriculture is absolutely not necessary and only causes harm and suffering for taste pleasure. At least with LLMs we get many positive advancements from them. I don't see these sorts of comments every time someone posts a burger review.

Indeed, analysis is a bit more loose in its arguments, and so I've found LLMs tend to make more mistakes there.

Nope. Codex formalizes much better than any tool with exception of Aristotle from Harmonic.

https://github.com/vjeranc/fixed-rtrt

M3 module was formalized fully purely from experimental data and from a nudge by earlier versions of codex in 15-30 minutes in a simple write/compile/fix-first-error loop. I was a bit surprised how fast it picked up the pattern but given there was a paper from '70s it became clear why later.

Undergraduate? No. We've had calculators able to solve undergraduate problems for decades. AI doesn't change the need to understand how calculus works any more than calculators did. The foundations remain valuable.

Graduate? Yes.

I don’t think it’s just mathematics. We don’t hear enough about this, but if I think back to my undergraduate years, which were less than 10 years ago, every homework assignment and every take-home exam I had would be trivial for LLMs to solve at this point I wonder what is actually happening on the ground.

May I remind you that this is Timothy Gowers. He says he doesn't understand, but he most certainly has far greater capacity than most to detect complete junk from a maybe plausible argument. His colleague is even better able to judge this, hence why he sent it to him.

Also if he did send me complete junk, I would still parse it for multiple days to see what is there.

This is beyond ridiculous to say considering whose blog this is.

For those that don't know, this is Timothy Gowers. He is one of the most accomplished mathematicians in the world. Like Terence Tao, he is considered one of the world leaders in mathematics and tends to have good judgement in where the field is going.

Even without that knowledge, no, this article is certainly not AI generated. It has none of the tells.

What makes you think either the tweet or blog post are AI generated?