Robot witch of @MissionRobotBay. Head of DevRel @NewTheoryAI. Often chased by tiny leopards. Prev: @0xPolygon, @Optimism
San Francisco, CA
Joined November 2008
- Tweets9.5K
- Following3.1K
- Followers6.1K
- Likes22.9K
Pinned Tweet
When I joined Polygon, one of the first things I wrote was this document. It's incredibly important to me, because it's about what's important to me. I share it with every new member who joins my team.
A User's Guide to Working with Mattie
Does it count as enfranchising robots if they cast a human vote?
We’d better start polling them just to be safe
Is this the first robot to ever vote?
I built an @OpenAI Plugin to guide me through our lovely props, and had a @MissionRobotBay arm mark my ballot for me : )
We can’t be far from a data broker purchasing TMZ so they can sell out-of-distribution car videos of, say, what happens when a Learjet missiles into the ground in front of you
In SF you can get a full V100 GPU cluster installed for free in exchange for letting the owner of the cluster live in the same room for as long as it’s there
Searching for Oakland date spot ideas with Gemini, and found two that are perfect, except Gemini made them up based on an obscure music collective and an alien species from the Cthulhu Mythos.
Badass Mission Robotics member that started by testing on our bots 👇
We just launched Odin Robotics (YC F26) 🦾
We deploy AI-powered robot arms that take over repetitive factory work: loading CNC machines, packing, palletizing.
Traditional automation means rebuilding your line around the robot, and it breaks when anything changes. Ours fit into the line you already have and handle the variation of a real factory floor.
Programmed control for the steps that must be exact, learned skills for everything that varies. Every deployment makes the next factory faster to set up.
If you run or know someone who has a factory with repetitive manual work, we'd love to talk → odin-robotics.com
The most interesting team I met this week was a weather-changing startup. They corrected me that no, not cloud seeding. They’re controversial in the cloud seeding community. (Also, apparently there is a cloud seeding community.)
They change weather patterns using gas flares natural gas companies ignite when they have excess capacity. When those companies have more supply than demand, they set the stream on fire rather than allow pressure to build up in the connected oil well. This startup finds wells that have regular flares and use that as the vehicle to change the surrounding weather chemistry.
They’re bullish on a Soviet-era experiment that showed you may be able to use the flare’s vortex to suck pollution from the sky.
Thank you @OpenAIDevs for having us over to talk about codex, robots, and @MissionRobotBay!
I'm incredibly lucky to get to work alongside devs doing everything from getting a robot to pour a Guinness to building glasses-free augmented reality to controlling 40K+ robots all around the world. Sometimes they even make the robots dance.
If you're making robots do surprising things, or giving AI a surprising form, come through 👇
Meet @Scav, who brings builders together in a shared robotics community where they push the edges on what's possible with AI in the physical world.
In one experiment, GPT-6 Astra remotely controlled a robot arm:
The crazy thing about generalizing models is if you buy a laundry folding robot, you also probably get a massage robot
While humanoid makers work to bring robots into our homes, Japan’s MW plans to sell homes with a housework robot hanging from the ceiling.
In MW’s design, MW bot travels along ceiling and wall rails, keeping its base off the floor. It uses two arms for chores, works alongside MW Cargo for transport and tucks away in a Nest when idle.
Before anyone moves in, MW can plan counter heights, laundry stations and travel routes around the robot’s reach. That can reduce the variation it has to handle.
MW demonstrated a prototype folding laundry in September, but it was teleoperated. The company is targeting an autonomous home demo in 2027 and sales of robot-equipped homes in 2028.
For new construction, this approach makes sense. The trade-off is the installation work and a working area defined by the rails.
Ok I love this. @samanfarid and @formic refuse to do proof of concepts. “They are a waste of time.” They get right to work and take balance sheet risk in order to go straight to the customer experience. “Your job is to make it as risk free as possible for them.”
Key questions so far tonight have been:
- How do you know if a robot company is real?
- If you had to deploy and be profitable in less than six months, how would you do it?
- Where… the heck… did you get that?
With @EmbodiedCoX and @Scobleizer
Was rooting around in a drawer and found this CD from 1994 by @OpenAIDevs. Anyone have a cd player?
btw you should come to this if you a) want to go very deep on deployments and also b) if you want to see me in a giant witch hat
luma.com/9qyybpyn
Mattie Fairchild retweeted
Thursday 10/29 (cont'd)
‣ Surreal Dinner Series - Recursive Self Improvement talk w/ Recursive Co-founder luma.com/surreal-8jm6
@veelarco @mercebent
‣ Founders, Funders, and Fresh Lime! luma.com/cgzl4lmz
@aspec00
‣ The Deployment Moat: A Physical AI Industry Panel luma.com/9qyybpyn
@rachelwrr1118 @skylar @Scav
‣ Text-to-Screech: Haunted TTS Launch Event luma.com/assemb-42oc
@assemblyai
‣ Pancake AI x Corgi luma.com/usecorgi-fda6
@usecorgi @jean_ette_li @erikalee @lauradang0 @tristan_cte @GuillaumeBasalt @francoisdefitte @unfairdotso
‣ Agentic AI Day: From Models to Systems That Act luma.com/cpqlpbpi
Mattie Fairchild retweeted
Robots are getting better at completing tasks. But what happens when they make a mistake?
We fine-tuned π₀.₅ on two real-world manipulation tasks to showcase whether recovery demonstrations can improve policy reliability more effectively than additional successful demonstrations. 🧵
I should call him
xitter.gq/scav/status/1751293264…
