Hacker News new | past | comments | ask | show | jobs | submit
> Our first formalization of Fermat’s Last Theorem, performed by an unreleased research version of Claude, took 11 days. It relied on a purpose-built harness along with guidance from a human researcher. By contrast, a team of 1,000 Fable 5.2 agents fully formalized the theorem in just 30 hours, with minimal human intervention. Fable 5.2 started from minimal prompts, built its own ways to coordinate, and used no mathematical references, only a list of intermediate theorem statements to score progress. It needed one-third of the action tokens as our first formalization and produced a proof seven times shorter. As can be seen in the figure below, Opus 5.5 (with the same 1,000-agent setup) completed only 26% of the formalization in 18 hours, and barely made any more progress after 30.

Most people don't really have intuitions on what it means to push the frontier now (because AI has surpassed everything they are good at, for example, coding). We are truly inside the singularity (using Kurzweil's definition, slightly paraphrased - the time when technology progresses so fast no one can keep track with it).

loading story #50432151
I think this is upside down.

AI is actually still really bad at architecture. It produces very good localized code, like an auto complete. It produces so much of it, that it can be overwhelming, but it's not remotely better than the equivalent aka an infinite number of talented humans doing the work.

The scale of it (aka 'make me a video game') is daunting and I think misinforming our intuition as to what is going on.

LLMs are brute forcing outputs and attacking math problems is not a singularity, it's just more compute.

What we are witnessing is the ability for machines to use fuzzy concepts, instead of very strict code, which allows it to map into a much greater variety of things, aka code, pdfs, web - and crunch/compute the content.

It's just compute.

Architectural goals just require more context than most people give an LLM. If you give that context to the LLM it will make a good effort at suggesting a solid architecture, and implement it for you.

Don't complain when it optimises some code for throughput when you haven't told it you're working in a low latency or realtime system

loading story #50437896
loading story #50434308
loading story #50436197
loading story #50430723
loading story #50431080
loading story #50433862
loading story #50433226
loading story #50430491
loading story #50430827
loading story #50432267
loading story #50432294
loading story #50432756
loading story #50431882
loading story #50430443
loading story #50430958
loading story #50435986
loading story #50431174
loading story #50432156
loading story #50435452
loading story #50434469
loading story #50438326
loading story #50430632
loading story #50430680
loading story #50432389
loading story #50430636