Anil Kumar

We're building a surgical robot capable of reaching any brain region. The goal: a generalized neural interface to help solve any condition that originates in the brain.

There are only two honest metrics when it comes to benchmarking intelligence: novelty and efficiency. You don't need intelligence to solve a known problem (only memory). And you don't need intelligence to solve a problem via brute force. But to solve a novel problem efficiently, intelligence is the only way.

i ate 6-12mg of xannys every single day for eleven years and quit cold turkey with no problems. this sounds like a skill issue unfortunately.

After 1,000,000+ engineering hours, we're ready to give an update on what we've been making and how. Construction is the largest, least productive industry on Earth, and it’s starting to affect housing. We're here to fix that. Thesis: We can build quality homes — any type — faster and cheaper than most of the industry through industrialization. Here's how: Construction's biggest problem is logistics: getting the right materials, people, and information in the right place at the right time. So we invented a new type of manufacturing, Mobile Micro-Factories — small, modular, local factories that make and prepare every part of a home, then deliver it in stages, last-mile, to the site ready to assemble with non-specialized labor. We’re built to scale fast — we can quickly copy our factories into hundreds of markets worldwide. This is a really hard problem to solve; we're going to get there. We'll walk through the nuances in more detail over the next few weeks. Full film dropping soon.

The creators of SWE-Bench just dropped a really simple new benchmark every LLM gets 0% on. ProgramBench asks: can models recreate real executable programs (ffmpeg, SQLite, ripgrep) from scratch with no internet? We are far from saturated on model quality.