https://teapawt.net/t.co/tzUMoyaH3x // executing leadership from the root
Joined April 2026
- Tweets4.7K
- Following56
- Followers167
- Likes7.8K
Pinned Tweet
Who am I and why should you follow👇
Most leadership advice is written by people who haven't touched production in a decade. I'm a systems architect and engineering leader writing about extreme technical debt, stoic leadership, and executing from the root.
I finally finished the Silo Trilogy by Hugh Howey, overall enjoyed it but the ending of Dust felt a little rushed and convenient.
What to read next? Hmmm.
The Dog Stars really gets off to a slow start, then burst into life about an hour in. The pacing and tone is very inconsistent.
These guys are fucking legends. Their replies in this thread are non-stop fatalities 😎
Introducing Arete-27B: Qwen3.8-27B, post-trained to reinforce Western values.
Chinese open models are currently the strongest available, but they ship with state-aligned behavior in the weights. We built an 803-prompt eval across seven value categories, including freedom of expression, democracy, and privacy, and measured how often a model adopts China’s party line.
After fine-tuning on roughly 53 million tokens, Arete-27B's propaganda rate fell from 22% to 1%, in line with the Western models we tested. Its Hermes Bench score shows no significant performance tradeoffs.
The values in an open model are one fine-tune away.
Our eval, weights, and training data are available now along with the full blog post: nousresearch.com/arete
Continuing to purge my search index today, mainly non-canonical links from the past before I put tight controls in place on indexing canonical only.
This isn't even the most confusing thing to me: that would be 🇺🇸 using ounces as a measurement of weight and liquid volume 💀
Lonely nerds in the Valley now realizing the best way to get laid is to put "works at an AI lab" on their LinkedIn.
After a deep analysis of the 81m web pages in the Greppr.org index yesterday, I realized there is ~10% of that index that is low-quality entries that can be purged.
Doing that purge today, wish me luck.
The desire for an AI to push you to start the next task while the current task is unfinished reminds me of Product Managers. Not cool.
This is so cool!
Asimov 1 is an open-source humanoid robot you can build, train, and customize.
- 1.2 m tall, 35 kg
- 25 powered joints
- Open hardware with off-the-shelf parts
- Open-source SDK and simulation models
Available as a DIY kit.
Order yours at menlo.ai/order
