>we developed a predictive model (R^2 = 0.513) that uses measurable task properties like tool count and decomposability to predict which architecture will perform best.
Is this going to be released for general use?
Towards a science of scaling agent systems: When and why agent systems work
41–42 of 42 posts
Re: Towards a science of scaling agent systems: When and why agent systems work
#42Earlier quoted context omitted.
The underlying models are impressive, be it Gemini (via direct API calls, vs the app or search), I would include alpha-go/fold/etc in that classification The products they build, where the agentic stuff is, is what I find unimpressive. The quality is low, the UX is bad, they are forced into every product. Two notable examples, search in GCloud, gemini-cli, antigravity (not theirs technically, $2B whitelabel deal with…
Antigravity is not a windsurf reskin, at least not today; it introduces many concepts and optimisations that you wouldn't find anywhere else, and in my workflows Gemini 3 Flash in Antigravity also happens to outmatch Claude Code with Opus 4.5 on some really gruesomely complicated tasks (i.e. involving compiler/decompiler work.) They are really cooking with Flash + Antigravity.
The Gemini family seems to be better at high-level, generalizing
The Claude family seems to be better at localized coding tasks
I'm a big fan of Claude Code based prompts with Gemini 3 Flash in my coding agent. I'm unwilling to use any new Google products at this point in time. Used to be a stan, they have pushed me away and I'll never be a stan again