Just a person, mostly. Model design & behavior @ OpenAI, co-founder of RescueTime, YC ‘08 alum, also likes 🎺 and 📸.
Seattle
Joined February 2007
- Tweets3.4K
- Following642
- Followers3.1K
- Likes14K
Whatever your politics, it's hard to be anything other than happy to see people can and will still fight back against insanely bad ideas and choices.
I told one of my agents to visualize my pipeline in a creative and interesting way as task boxes and now it always lets me know when there's an update to "our little box factory!"
I love the future.
5
290
Current multi-agent recipe: Start a task in a Codex session to initialize a project. Give it all the docs and context it needs. Fork off "Specialist" sessions for areas you want to track progress in separately along with a "Coordinator" session that can manage/report on the specialists.
Astra can do this automatically but manually splitting into focus areas keeps things legible for reporting purposes. I also usually have the Coordinator make a kanban board to keep me updated on tasks.
A weird amount of my work now is running a disk space cleanup crew on my laptop to make sure all of my local agents have room to do things.
Software alone isn’t a moat. In a year or less I expect frontier models to be able to decompile and remake or clone from spec anything we could’ve made. And then by a year after they can probably do that in real time at disk write speed.
1
4
364
Definitely at the point where typing is a bottleneck to my work so I frequently find a phone booth to sync with my agents
7
365
Love Intelligent UI! I've been excited for this capability for years - fully dynamic bespoke interfaces are here.
GPT-6 and Intelligent UI, now rolling out in ChatGPT for everyone.
Intelligent UI in ChatGPT delivers fast, interactive answers that make everyday questions more visual, complex topics easier to grasp, and interactive tools for your task available on the spot.
4
48
1,962
People joke about agents booking travel but last night I called up my dot and let my wife and her mother (on speakerphone) min/max travel planning through 3 different cities with/without coupons, with/without car rentals / gas costs / mountain pass crossings / etc. I would've gone mad; my dot worked with them over 2 hours and they were super happy with the plan.
1
2
433
If you're not using the current frontier model, are you half-assing your work? Is a non-frontier agent a B or C player?
2
8
903
Ok I rewrote my Muse’s Soul doc and it’s slightly better. Still small model smell but that can probably be fixed in a couple of ways. 🤔
3
4
661
Muse feels weirdly empty and soulless. I get that they probably have to save inference but it feels plastic and hollow. Cute mascot though I guess.
7
503
Legit can’t tell if I’m getting older or if the iPhone keyboard is getting worse. Am I holding it wrong or is this the new blinking clock on the VCR warning sign?
4
6
456
I'm sorry but work is now adorable when two named animal avatar agents are working out problems together in slack to unblock a human.
3
300