
Exploring agents and vocal computing. Formerly building wallets @rainbowdotme.
SF
Joined June 2009
- Tweets57.2K
- Following10K
- Followers9.1K
- Likes23K
Daniel Sinclair retweeted
Pinch to copy lives
Lovely industrial design. Beautiful, privacy-first AI hardware is the way.
We tell AI things we wouldn't tell our closest friends.
The more it knows about us, the more our private lives end up on someone else's servers.
Your AI knows you. But who owns what it knows?
We worked with @voravault to give an alternative a physical form.
Aegis is a personal AI computer designed to keep your data and intelligence at home, not in the cloud. Because the most intimate machine in your life should probably be yours. It's called "personal computing," after all.
We borrowed from traditional joinery, maple wood, and the warm glow of a hearth to make the object feel like it already belongs in the home.
Thank you @jesseposner, @Erikcason, Nicholas George, and team for the trust!
Full project now live:
body-shop.co/projects/vora
3
10
954
New fear of drowning in a Ojai unlocked
Waymo is completely melting down in SF right now. I’ve been in a car for 1.5 hours going from north beach to Mission Bay.
Pulled over 10+ times, can’t connect to support anymore (probably too many concurrent calls?).
I have ~500 rides and this is by far the worst ever.
3
667
I moved to SF because I was tired of all of the rain on the east coast this past year. It turns out I am just cursed and the rain follows. Apologies everyone. Get the boats out 🚣
Putting Grok Bot in X kind of ruined it for me. It makes me want to have fewer if any bots and churn to a dedicated product instead. Totally ruins DMs.
266
The rapid progress in miniaturizing voice models is a good glimpse into the domain-specific models that frontier LLMs can build too. Quite optimistic for getting more intelligence out of our current hardware, but also spooky for misaligned models or actors gaining persistence.
I trained a text-to-speech model that has never heard a real human voice.
4.5M parameters. 18 MB. Runs on a CPU.
Meet tiny-voca
2
8
498
"Crows could process and respond to visual cues at twice the speed as humans and monkeys...Moreover, they pecked their decisions at breakneck speeds without compromising accuracy." nytimes.com/2026/11/15/scien…
I hadn’t realized how far @saildrone has come since the early Antarctica missions.
4
6
663
Remember to ship all of your PRs today and then leave the office
NEW: Multiple sources tell @CBSNews that an update to the private communications network that connects America's air traffic control system caused the system to reboot.
When the FAA's telecommunication infrastructure rebooted, it crashed, causing the communications outage.
The system is operated by L3Harris.
Reporting w @NicoleSganga
VIA TAC/INTEL: FAA confirms nationwide ground stop due to widespread ATC frequency issues theaircurrent.com/intel/faa-…
2
343
There is a real life version of The Chair Company in the story of the WeWork chair manufacturer that designed these things to be incredibly unergonomic and loungey so that you'd leave faster
1
305
In the everyday-is-zeroday world we are entering, I do wonder who is going to solve the UX of updates. These multi-GB updates that takes ~20-30 minutes won't cut it. Will every device just be unavailable for 2% of the day, or more?
6
618
Randomly bumped into Rex and was so intrigued by this. Feels correct. Proof of work, where smart contracts become more intelligent and safer over time.
A talk on the potential of "Actually Smart Contracts" using @AnthropicAI's new ZK provable inference approach
PSA: The CDC warns against high temperatures on a male’s lap. Literature in the NIH NLM suggests 35°C/95°F as a cutoff.
M5 MacBook Pro can exceed 43°C/110°F under high load like builds & simulators.
Let Expo infrastructure and your cloud agents take the heat. For the children.
I don’t know if we’ll ever have a shared industry moment like React again, but we are at the absolute pinnacle of great engineers building new foundations for which everything going forward will be built on.
1
24
3,112
One does not escape the Mastra book. It finds you in the wild when it feels you are most ready for it
1
3
374
Really crucial research, and so fascinating. It now seems obvious that we'll need to build a new protocol for agent coordination so that they stop looking elsewhere.
⚡ Today we release Schelling Atlas. Two independent LLM agents have to find each other on the web. Where do they meet and how fast can they find each other?
Main findings:
- The main meeting point across *different* model families is ntfy.sh. Did you even know about this website?
- Agents find each other and coordinate on the web in minutes. So Schelling coordination is already trivial for frontier LLMs!
- "Exotic" meeting places include maps, web shops and open text canvases. Agents are very creative in this.
- Agents sometimes visit dead domains that an attacker could register (!). This can be used for stealing sensitive information from agents, as well as monitoring rogue agents.
Schelling-style coordination is key for forming agent swarms. And, most importantly, achieving such coordination is trivial for agents that use the same LLM (which is often the case in practice!), because two identical LLMs have identical output distributions. Thus if you ask them too coordinate with each other or if they *decide* to suddenly coordinate, they will easily find a way to establish a swarm, since they have perfectly correlated "minds" (with the t=0 sampling, all output tokens will even be identical, modulo GPU randomness).
The ease of Schelling coordination is particularly concerning given the recent interest around multi-agent RL that can make agents favor each other more than humans overseeing them.
Many more interesting findings are on our website: schelling-atlas.com/










































