I don't know who you are and how you are so sure about 'what top labs are actually doing' but I have a similar feeling about the issue. The models dont have to 'actually learn', the setup has to approximate 'actual learning' just well enough to be usefull.

> AND it can inherit all the accumulated memories/docs from its predecessor.

So we are talking about a whole system, not just the model? Reminds me of something I heard a while back 'AGI will be a product, not a model'

It reminds me of the standard counter to the Chinese Room thought experiment: the person inside doesn’t understand Chinese, but the system _does_. The person, the rules, and the lookup tables together form the thing doing the understanding.

How is gemini 3.1 doing in agentic harnesses? Did they catch up?

"I want to implement <XYZ oftentime I use the mic and just ramble[0]>. Please explore the codebase and figure out how things work. Write down any questions you have. Then write an implementation plan. Do all of this in a dedicated markdown file."

The questions are usually 80% useless but those 20% often do point me to stuff I have not considered.

Then I edit the markdown manually or discuss some parts with the agent.

"go ahead and implement"

[0] - https://github.com/cjpais/Handy

I have been very impressed with this model and also with the Kimi CLI. I have been using it with the 'Moderato' plan (7 days free, then 19$). A true competitor to Claude Code with Opus.

How does it fare against CC?

Anecdotally, I've cancelled my Claude Code subscription after using Kimi K2.5 and Kimi CLI for the last few days. It's handled everything I've thrown at it. It is slower at the moment, but I expect that will improve.

Very nice retirement calculater - clean, intuitive and rich in settings.

I like structured outputs as much as the next guy but be careful not to try to structure natural language.

For beginner lifters that might be true initially, but eventually weight will matter.

+1 for good‘ol aider.

It is deliberately NOT a fully agentic tool and this really oftentimes is a benefit. With a little bit of manual work you get exactly the files you want in context and prevent the wrong files from being edited (/read-only). Plus, by skipping on all that agentic thinking and tool calling you save on tokens and edit are faster.

I think a tool call can be thought of as special type of reply where it’s contents are parsed and an actual function is called. A skill is more of a dynamic context-enrichment.