Figma made design multiplayer. We made coding agents multiplayer.

Introducing Superconductor: put your team and your agents in the same session.

Run Claude, Codex, Pi, etc in the cloud, see live app previews, and ship from any device.

Give it a spin at superconductor.com

We evaluated Claude Tag, Dust, QM, Viktor, and Superconductor on a rubric based on the Multiplayer AI Manifesto principles.

You can set which criteria matter to you and see which one is best!

Let me know which platform we should evaluate next.

multiplayer-ai.com/compare-platfo…

Having trouble keeping track of all your favorite agents and models?

Check out our new agent picker:
• search across every agent
• effort levels right on the row
• star your favorites
• quickly launch multiple agents on the same ticket

superconductor.com/blog/agent-pic…

is this about right?

You can now launch Superconductor agents directly from @linear!

Assign any Linear issue to Superconductor, run coding agents in the cloud, see a live preview once the agents are done.

superconductor.com/blog/linear-in…

A perpetual problem: the more visible activity there is, the more impressive the work looks. Ten years ago people were building unnecessary distributed systems, with >1:1 microservice:dev ratio, that could've been simple monoliths. The same thing is happening with "factories".

The later it gets, the more baroque my Opus 5.5 prompts become. It is a marvel of a model, but it's a bit of a diva (at least in my mind). This is why demand for frontier tokens is currently essentially infinite. The more capable the AI's get, the more insane tasks we give them.

Your coding agent read all 48,213 lines of your repo.

It still has no idea which bug you mean.

The bug is in Sentry. The ticket is in Linear. The design is in Figma.

Now, add Connectors on Superconductor once for your workspace, and agents go get the context themselves.
c

Claude: earth-toned, dimly lit, cozy sweater named. Autumn coded. Humanpunk.

ChatGPT, Gemini, and Grok: colorless, concise, inhuman names. Winter coded. Cyberpunk.

This leaves a giant opportunity for an AI lab that is Spring/Summer coded. Solarpunk.

New in Superconductor last week:

- Google Workspace connectors for Gmail, Calendar, Contacts, and Drive
- OAuth option for many of our connectors
- A redesigned agent picker
- Sonnet 5.5 + GPT-6.1 Sol

Full changelog:
superconductor.com/blog/weekly-ch…

> I'm making a diagram explaining a product feature. What teammate names should I use in the diagram?

GPT: Maya, Alex, Jordan

Claude: Maya, Jordan, Priya

Gemini: Alex, Maya, Sam

Grok: Alex, Jordan, and Sam

---

This is what is happening to EVERYTHING.

AGI has been achieved externally

Cells in your body do not know the goals or even the shape of you, the organism they comprise.

And you do not know the goals or shape of the organism you comprise.

Okay, you’re not reading code anymore. What ARE you reading? Teen Vogue?

Who stops a bad guy with a misaligned superintelligent AI?

A good guy with a misaligned superintelligent AI.

dude what is going on with google man. they sunset gemini cli, which was never good, replace it with antigravity, which i really don't understand what it even is -- is it for coding? is it a cowork-style assistant? what is the antigravity ide and how is it different than the open antigravity app, update it, and get this. why would my name be alice? why does it have memories -- this is the first time i used it!

Hey so, like, the frontier models-in-training are clearly not fully aligned. Why are we training them?

@tszzl please explain like I’m 12 or whatever

What does it mean to work with AI agents in a fundamentally multiplayer way?

I can review a pull request by chatting with the EXACT SAME agent that was used to develop it (which cost almost $700 lol).

I can see the code diff right there, and leave comments to the agent on it.
l in the human developer with an @-mention.

And I never have to check code out locally. The live app preview is running on the agent's sandbox, so I can use it if I don't trust the agent's (or the QA agent's) screenshots/videos of the feature.