AI as a Research Partner: Antigravity Rolls Out Teamwork Updates
Source Google Antigravity blog

At a glance
Google has updated Teamwork in Antigravity. In Teamwork, several AI agents suggest ideas, critique them, and improve one another's work on their own for hours or even days, and it can be used now as /teamwork-preview on all paid plans. The update builds on the launch of Gemini 3.7 Flash.
Key points
It solved seven open problems in math and theoretical computer science, and Knuth's Cycles Conjecture was confirmed with the Lean proof checker.
It scored 71% on TCSBench, a theoretical computer science benchmark, the highest result in Google's internal testing.
It built a RISC-V CPU simulator that boots an operating system from scratch and contributed speedups to open-source libraries such as Eigen.
Gemini analyzes each request and automatically picks a fitting team setup, called a pattern.
Why it matters
By having agents check each other, it reduces the problem of loosely organized agents following one another's early mistakes, so AI can take on large, open-ended research and engineering problems.
Original article
Teamwork: when AI becomes a research partner
Google Antigravity blog · Aug 27, 2026, 7:00 PM GMT+9 · antigravity.google
Read on Google Antigravity blogAI read the original article and summarized it in plain language. It may contain mistakes, so please check the original for accuracy. The article's copyright belongs to Google Antigravity blog.