Most errors are probably responses that didn’t finish before their 3K token limit. They’ve measured how well RL is able to shorten the response to their limit.

In a similar way that Google Maps shows eco routes, it’d be fun for them to show “safest” routes which avoid areas with common crashes. (Not always possible, but valuable knowledge when it is.)

For example, only 7% of pharmaceutical research is publicly accessible without paying. See https://pmc.ncbi.nlm.nih.gov/articles/PMC7048123/

The progression from basic arithmetic, to complex ratios and basic algebra, graphing, geometry, trig, calculus, linear algebra, differential equations… all along the way, there are calculators that can help students (wolfram alpha basically). When they get to theory, proofs, etc… historically, thats where the calculator ended, but now there’s LLMs… it feels like the levels of abstractions without a “calculator” are running out.

The compiler was the “calculator” abstraction of programming, and it seems like the high-level languages now have LLMs to convert NLP to code as a sort of compiler. Especially with the explicitly stated goal of LLM companies to create the “software singularity”, I’d be interested to hear the rationale for abstractions in CS which will remain off limits to LLMs.

It is a simple Codex meta-"skill" that uses a CLI to analyze your past Codex sessions and generate reflections. It uses headless Codex subagents running in parallel on temporary copies of conversation history. These reflections are orchestrated by the CLI and a Codex agent can synthesize them for downstream use. Reflections are cached for efficiency.

Repo: https://github.com/olliepro/Codex-Reflect-Skill

Comments URL: https://news.ycombinator.com/item?id=46738461

Points: 3

# Comments: 2