so its been a couple days since we started building a game in public, have you done something yet?
make a change.
whim.run/demo
"At Anthropic, almost 100% of our engineers are running 100+ agents"
We're providing a free week of compute for you to do this! Try it out whim.run
It's hard to run and realistically manage 100+ agents at the same time on your computer.
But we make it super easy on Whim to run everything on the cloud so that you can scale and access them anywhere, anytime.
Creator of Claude Code:
"At Anthropic, almost 100% of our engineers are running 100+ agents with self-improving loops
self-improving loops help agents become better with each run."
in a 1-hour podcast, Boris explains how they build agents loops from sratch.
Claude + loops + routines + dynamic workflows - that’s the secret.
Watch the talk, then read how to apply the same playbook to quant trading below.
What if a hundred of us built a game together?
Let's build a game together.
whim.run/demo?utm_source=x&u…
We spun up a shared workspace and dropped a basic Asteroids game inside. From here it's yours. Add what you want, break it, rebuild it.
whim = claude code / codex + steroids
stop running 2-3 agents on your laptop and orchestrate 100s on the cloud.
don't wait. free compute only during alpha!
Cursor, Claude Code, Codex are all single-player. Whim is multiplayer.
Whim is a shared workspace where humans, agents, code, context, compute, and history live together.
teapaw.org/whimdotrun/status/2056…
But the deeper problem is collaboration in a team.
You build hours of context with your AI. Then a teammate picks up the task and their Claude starts from zero. Every handoff is a reset. And because agents let everyone move so much faster, nobody has visibility into what the rest of the team is even working on anymore. You're moving at AI speed with zero visibility.
Every major leap in human capability came from a leap in collaboration. Language let us coordinate. Writing let us remember. The Internet let us think together and organize at scale. It was not about getting smarter individually, but about sharing context and collaborating.
AI agents are powerful but exist in a vacuum. Each one starts from zero. Each one works in isolation. Each one forgets the moment the session ends.
We're building the language that can help agents collaborate with each other and with us, as a real team.