Join the movement and participate on Uniswap.
Participate
People at NACS asked what the robot does. Inventory? Restocking?
This is the new store manager.
They looked confused. Good.
We were a couple of days late — NACS, then San Francisco, then a Friday keynote. The conference is the world's largest for convenience stores. We showed Cactus, an AI that runs the store, and robots plugged into it.
Retailers name labor and rent as the biggest cost. Most of them still don't know what's going on inside the store.
A Chinese retailer spent the last couple of years taking human decisions off the floor — what to carry, how to price, anything a person used to decide. This year they published the flagship result. People at the booth guessed 5% to 60%. The number was 258%. That's their case study, not ours. The point we made with it: the frontline needs brains more than muscle.
The robot keeps a digital twin so the AI can tell you how to run the store. Worst-performing fixture by shelf meter. What to do about it.
New leads cover well over 20,000 addressable stores. One of those accounts is over 20,000 on its own. Another ~800, another ~300, another ~1,000 — brands you'd know. They want to get to pilot fast. Not only robots. People want to pilot Cactus with phones.
Build a digital twin of the store. Talk to an AI about how to run it. That's the product even before a fleet shows up.
Retail tech companies were in the mix too.
We're in San Francisco for the Humanoid Hub keynote. Friday, 11:20 local, free: humanoidhubconference.com.
Part of the demo: a BracketBot pointing at a phone. Blind. Not tracking the phone with its cameras. It knows where the phone is because we built collaborative proprioception — the sense that lets you touch your nose with your eyes closed, except across robots and phones. The robot knows where every other robot is, and where every phone is, without looking.
We'll keep putting previews on Discord and X through the week.
Arshak has glasses for retail staff. Stand in front of a shelf, see that shelf. Walk to cookies, see cookies.
The glasses aren't doing spatial computing. These displays wouldn't send video, so we couldn't run SLAM on them — not even remotely. He used the attention-tracking setup from the lab: CCTV figures out where the glasses are looking, and we send the right overlay. Display glasses. The cameras do the locating.
He also built a spatial video editor on the real world web.
Auki is making the physical world accessible to AI by building the real world web: a way for robots and digital devices like smart glasses and phones to browse, navigate, and search physical locations.
70% of the world economy is still tied to physical locations and labor, so making the physical world accessible to AI represents a 3X increase in the TAM of AI in general. Auki's goal is to become the decentralized nervous system of AI in the physical world, providing collaborative spatial reasoning for the next 100bn devices on Earth and beyond.
X | Discord | LinkedIn | YouTube | Whitepaper | auki.com
