Traves retweeted
just beat kaparthys nanochat benchmark (on a single h100 for now)
pretty happy with this, as it's using alpha a language/compiler i've been working on for four months, up against his autoresearch'ed llm.c+cuda+cublass implementation
Traves retweeted
kind of like this remotion ai generated video, also a sneak peak what our team is building, let me know if you math-adjacent and want to have a play (i know nothing about math)
Traves retweeted
claude + blender helping me with design inspiration to make a woomera (spear launcer) from a piece of ironwood tree i cut down to the core
Traves retweeted
Deploying DiffusionGemma-Jev (djev) just got a lot easier. You can now spin up a Jev API-compatible endpoint on Google Cloud Run using a single command.
Performance is solid: ~35-60 ms for single step latency and batch@32 is ~100-123 requests/sec.
It's a straightforward way to experiment without needing your own GPU. Runs at roughly $3/hr and drops to $0 when idle.
Get the code and instructions here: github.com/taeold/djev-run
RT @ajaxdavis: stop supporting this horrible behavior from a company and its employees
it's become completely dishonest
changing the natu…
This quoted post is unavailable.
decided to switch from Claude to codex for a month, blew through my weekly codex usage in 2 days... what gives? i didnt even hit the hourly once. @OpenAI @OpenAIDevs
Traves retweeted
CRAP: Conditional Resource Access Protocol
ajaxdavis.dev/crap
Yo anyone else’s weekly usage on @claudeai Claude code just seem to be going super fast in relation to your 5 hour window somthing seems off. @ClaudeDevs
Traves retweeted
In the next version of Claude Code: subagents run in the background by default, so you can keep talking to Claude while your subagents work
If you want your agent to run in the foreground, just tell Claude
Traves retweeted
As engineering, product, design, DS, etc. melt into a new kind of role, I was reflecting on what roles might look like in the future. For example, when I look at the Claude Code team I see what I think is five archetypes:
1. Prototyper: comes up with brand new ideas; churns out many ideas, most of which don't ship
2. Builder: quickly turns a prototype/idea into production-grade product/infra
3. Sweeper: cleans up the UI, simplifies the code and system, unships, optimizes performance
4. Grower: takes a product that has been built and iterates on it to improve Product-Market Fit
5. Maintainer: owns a mature system to make it secure, reliable, fast, and efficient as it scales
Many people span across 2 roles, and sometimes 3 roles. I also notice that these roles are not really tied to job function -- eg. across Anthropic, some designers match category 1, some 2, some 3; same for engineers, PM, DS.
A healthy team needs a mix of these, depending on the product:
- A product that is new and pre-PMF needs people that are strong at 1+2+3
- A product that is growing and has found PMF needs 2+3+4 and some 5
- A product that has strong PMF needs 3+4+5 and some 2
Maybe product roles of the future will look more like this, and less like the domain-specific roles of today?
Traves retweeted
We're open sourcing a 9B model that extracts structured data from documents at near-frontier performance.
- 90.2% on our bench, vs Gemini 3.5 Flash at 91.3%
- Leads extraction models like NuExtract3 (81.5%)
- 9.5s p50 timings
- Pass JSON schema
Traves retweeted
Replying to @garrytan
depends on the task, some require more guardrails than others
for exploratory tasks, absolutely, but a loop can be executed many different ways:
- user driven
- agent driven
- workflow driven
- cron timer driven
- goal driven
loops can be used w all of those easily














