The Erdős problems have turned out to be largely brute force or finding older results.

The Feb 2026 GPT-5.2 theoretical physics paper was a result of “dialogue between physicists and LLMs”, called “grad student level” by experts in the field, used a “custom harnessed” “internal OpenAI” model with “20 hours of reasoning”. Quotes from OpenAI blog.

The Matthew Schwartz physics paper with Claude this March involved “51,248 messages across 270 sessions, producing over 110 draft versions and consuming 36 million tokens”, and the actual contribution was Schwartz finding an error in Claude’s solution.

I would have personally gone for 75%, 85% and 95%, which are all still best case scenario answers.

Had I taken on chatbot advice on electronics or chemistry I’d have died every couple of weeks (doing some hands-on real world R&D in my basement as a distraction from software).

Plane of words: broadly correct. Everything is flattened to tokens and token sequences, and the training data is dominated by text tokens.

Reasoning: CoT tokens are mostly just tokens, more appropriately called intermediate tokens, and are largely disconnected from the end result. Including them improves the end result (user satisfaction), but does not imply reasoning. See for example Turpin 2023, Mirzadeh 2024, Pournemat 2025, Palod 2025.

Synthesising evidence: You can achieve SOTA summaries with LLMs, but this involves, for example, using a harness to generate dozens of summaries with different models, separately using some kind of vector embedding model to compare results to the original, and selecting the best match. This is not how most people are using LLMs for summaries. While this is being slowly RLVR’d in post-training, a one-shot naive summary underperforms more complex methods significantly.

Video thumbnails are a different beast altogether. And you might want to double check your assumptions about security considerations. If any of your ffmpeg, opencv, pyscenedetect code is running on your server, it might well be exploitable.

And on the ROI side, trying things out regularly, I haven’t seen the positive ROI in the limited time I’ve dedicated to exploring the tools. I’ve restricted experimenting to 4 hours per month, because spending more than 2.5% of the month chasing productivity improvements that realistically seem to be 10-20%, will quickly eat into those gains. After accounting for token costs, it ends up being a wash.

And stepping through those entries isn’t like browsing a modern local-first app [1], where you will just scroll through dozens of entries in milliseconds. It’s not like the slightly older and slightly slower Gmail interface. You’re clicking on each record and waiting 400ms-3s for it to load, as if instead of a 25Gb fiber connection you’re on dialup requesting the record from Epic’s headquarters in the US and proxying them via Australia.

[1] https://bugs.rocicorp.dev/p/roci

What happened to lights off, door closed, do whatever you want in complete darkness in the bed that you aren’t able to climb out of?

With any kind of creative work for hire, from architecture to advertising, from jingles to commisioned sculptures, the client’s taste and budget, more than almost anything else, determine the outcome.

Take Cannes Lions as an example of a competition and awards ceremony that essentially exists to define what ’good taste’ means within that industry. The client’s team is prominently credited alongside the creative agency. They get to climb onto the stage for the speech and they have a voice on whatever video clip is made about the project.

Partly this is to encourage more ambitious and spendy work for the industry at large. But everyone involved certainly knows, that the same creative team, with the same creative idea, could have ended up making something much worse working with a different client team.

I can’t stand AI slop, yet I think I’ve unintentionally argued in favour of people creating it, as long as it’s… good by some measure?

Anthropic did retire an interview take-home assignment involving optimising inference on exotic hardware, because Claude could one shot a solution, but that was clearly a whiteboard hypothetical instead of a real system with warts, issues and nuance.

Some of the best known laws from the ~1700BC Babylonian legal text, The Code of Hammurabi, are laws 228-233, which deal with building regulations.

229. If a builder builds a house for a man and does not make its construction firm, and the house which he has built collapses and causes the death of the owner of the house, that builder shall be put to death.

230. If it causes the death of the son of the owner of the house, they shall put to death a son of that builder.

233. If a builder constructs a house for a man but does not make it conform to specifications so that a wall then buckles, that builder shall make that wall sound using his silver (at his own expense).

That doesn’t sound like ceilings never disintegrated!