Hustling in SF | @meta MSL
San Francisco, CA
Joined August 2015
- Tweets1.3K
- Following2.7K
- Followers1K
- Likes4.4K
🫡
the most underrated part of muse is the model that powers it. thank you to the MSL research team—some of the most technically brilliant, hardworking scientists and engineers i’ve ever known. you brought original ideas to difficult problems, invented new methods, and battled through enormous technical complexity to build the jewel at the heart of muse.
thank you to the teams working on agentic post-training, tool and computer use, memory, personalization, alignment, safety, and multimodal understanding. thank you to the people building data, evals, and infrastructure beneath it all. so much scientific ingenuity lives in those details: designing a better experiment, finding the flaw in an apparently good result, staying with a stubborn problem until a new way forward becomes clear.
and now that jewel powers a beautiful product your grandma can use without knowing a thing about the extraordinary complexity underneath. it will just work for her. all those discoveries, all that patient work, becoming something simple and useful in another person’s hands. there is something magical about that. i’m grateful to be building with you.
2
156
Muse Feed really growing with me as I used it more and more
Pro tips: you can even customize the instructions for the feed 🔥
fun fact: the muse feed was my idea (one of the few things that was) after realizing agents with strong memory recommend content to you in a very different way than existing recommendation systems
they’re much closer to a good friend sending you things they think you might like based on what they knows about you, vs a slew of content related to what you just clicked.
it’s just the beginning here, but very excited for people to try their muse feeds and see what they do with it!
150
Current “AGI” is a bit weird to me
Like we are witnessing a digital intelligence lifeform growing as we speak
Right now is a babe or kindergarten, the intelligence and reasoning are just mimicking smart adults
1
2
280
Introducing Harmonici - A Private-by-default finance app for couples.
As a couple, we manage our finance separately. Every month, we pull our expenses into a spreadsheet and settle the differences.
No more spreadsheet, try Harmonici
harmonici.vercel.app/
#builtwithgstack
1
2
204
This is what I build during the @ycombinator Hack The Stackathon
Manual research on all the electronic components on a BOM sheet is tedious
BOM Guardian enriches the component's information for you, automatically
2
3
176
- We use @firecrawl to fetch info from distributor or manufacture website
- @reductoai to parse long PDF pages
- Embedded and stored in @mongodb vector db
- @openrouter with RAG for chat interface
- @resend to send summary email
1
147
Talking to an AI engineer be like:
We need a better harness to test this harness so we can harness the most out of this harness
60
For the few, AGI is already here
when token per watt continue to increase
when agentic UX improves
It’s like the calm before the Tsunami
56
Claude code has completely changed how I work
I find myself get pull into using bash command or CLI as I tried to work like Claude, think like Claude
Most UI are redundant, built for human, with micro-frictions to navigate, and it’s extremely inefficient
Bash is all you need
60
Saw lots of stories about solo founders leverage AI and making profits by just simply prompt their way to production SaaS
now I can 100% proven that’s true
just delivered my first milestone to a client that’s willing to pay for revamping their legacy system
1
98
I was worried about the future maintenance if its a burden but later I realized AI is just going to get better
I only need to setup a system, where, a user ticket in, a PR by AI out. I believe that’s totally doable today
1
79