The premature optimization of splitting animation work onto another process has done lots of harm to iOS (and web). For a brief while things were slightly smoother, but in exchange we ended up with uninterruptible animations and layout glitches from introducing unsolvable concurrency problems, plus various patches on top that ultimately make the system actually slower than if we just did things cleanly on main thread (or had access to proper multithreading in userland). Most tasteful UX cannot be decoupled in a computer sciency way
Smooth animations in general are great but calcifying the rendering pipeline for a decade is a huge cost
Remember how back in the crypto days folks asked "why can't we make these GPUs run useful work" and the typical answer is "making GPUs run useful work for money would defeat the purpose" and it turns out that's completely wrong in an important way
the flashlight method of comparing two projections of the same embeddings 🔦
2M points in webgl, color by spread/compaction, umaps aligned by procrustes.
I've had fun visualizing ml concepts with manim. Here’s a small two-layer MLP, where each layer computes h = ReLU(Wx+b).
This viz lets you see that every connection is a single weight value.
@danluu one of the big early ideas behind tailscale was a single computer is powerful enough to solve almost every problem, and that this is not widely understood. we need fewer distributed systems and more spreadsheet implementations.
Reverse Greenspun's 10th rule*:
Any sufficiently complicated program contains an ad hoc, informally-specified, bug-ridden, slow implementation of a neural net
* chenglou's 10th rule
It's time for a South Bay renaissance
Might be hard to recall, but used to be not that long ago that the center of Silicon Valley was squarely in Palo Alto, not SF. When I came to the Bay in 2010, that's where I lived and that's where I stayed and settled down to build my startups
It's a beautiful place to live, maybe the best weather in the world, teeming with top talent, safe and family friendly.... but the locus of tech has moved to SF., leaving the startup scene down here somewhat bare and deprived. Most of us are too busy to make it to SF more than 1-2x a month.
I'm committed to revitalizing the South Bay for startups - with events, dinners, social activities (incl parent/kids activities), etc. We can build a vibrant hub down here if a few of us make it happen.
If you're with me and want to be a part of this renaissance, reply or DM me and let's make it happen
I believe Freerange might be one of the biggest open-source projects written with Fable & Sol. And technically barely feasible with current AI
At its peak it had 50k lines of code. Was a long journey iterating & reducing it down to 18k today. The pivots were brutal
If I started it a few months earlier with worse models I'd have lost my mind. Thank you OpenAI for the internal model access!
My dear UI developers, ML practitioners, and fans of programming language:
Many months & billions of tokens later, I’m proud to present to you the first step in our long, long collective journey to turn vibe coding onto proof engineering, starting with: making user interfaces verifiable.
Introducing: Freerange, a zero-API tool that automatically deduces your code’s numerical ranges. By doing so, Freerange is able to prove that e.g.:
- your TS layouts obey your specified sizing
- that they’re free of NaNs and Infinity
- that your array indices stay within bounds
All of that, done statically. No browser, no running code, droppable into any codebase, and for the ML folks: RL-friendly