The peril of laziness lost bcantrill.dtrace.org/2026/04/12/t...

Job posting at Mercer for STEM Professor at $85/hour

Yeesh

arxiv.org/abs/2604.05091

Good challenge - let me verify rather than speculate.

If I wanted to pull an answer out my ass, I could do that myself

Introducing Prefab 🎨

A generative UI framework for building MCP Apps, data dashboards, or whatever you (or your agent) want.

In Python.

(Really.)

100+ shadcn components. Real React. No JavaScript required. And built right into FastMCP 3.2.

prefab.prefect.io

The Waymo Rule for AI-Generated Code AI-generated code doesn't have to be perfect; it just has to be better than you.

New post! I used to think the future was formal specs + model checkers. Now I think it’s English + tests + quality gates.

Free Disk Space for Developers Tools for cleaning up your system and freeing up space, especially for developers.

Shoutout to @tdhopper.com 's page that I still visit all the time
freespace.tdhopper.com

This is rad rbischof.github.io/windinet_web/

Fair point. Let me check the actual row count

sometimes CC reminds me of a new researcher who hasn't been burned enough to be skeptical of everything

Finally

the swear regex is so funny because all consumer-facing ML products have some variation of the regex filter list ahead of the model, it’s probably one of the most normal and relatable things I’ve read about AI labs

I wrote something: sockpuppet.org/blog/2026/03...

I’ve lived in nyc for 17 years and just cannot bring myself to say on line. Eternally grating

Grounding lets vision-language models do more than describe—they can point to where a robot should grasp, which button to click, or which object to track across video frames.

Today we're releasing MolmoPoint, a better way for models to point. 🧵

What is taste if not a better prior?

I wonder if the deep analytics, causal inference-y data scientists are going to have the last laugh now. Their less-measurable skills are less verifiable. They now have access to powerful engineering tools. And their taste / mental models / decision frameworks will likely increase in value. (2/2)

I feel like a lot of early data scientists got pushed towards analytics engineering or machine learning engineering (myself the latter) due to the need to "productionize" their work and probably because engineering skills are more easily measurable. (1/2)

Fake Jack tweet that says "we're firing half the company"

So first, this is almost exactly what the Gmail AI summary of the email he sent said and my reaction was "Lol the AI fucked up!"

The AI did not fuck up... 🫠

Shoutout to “DJuboor” for being the first public contributor to confingy! Version 0.1.1 is out now with pathlib serialization support and some small bug fixes

github.com/runwayml/con...

Have a shpritz!

A subway ad for ketamc.com

So we’re just doing ketamine subway ads now?

congrats!

Nothing makes me feel older than working on a random silly side project that absolutely would have crushed it on 2018 data twitter and yet is now so wildly out of step with the moment

Hi Ethan Nannernest caught my eye, specifically how you used computational geometry to optimize banana coverage on peanut butter sandwiches. The way you approached that optimization problem shows you think about algorithmic efficiency in practical applications. I'm working with a seed stage company backed by CRV, XYZ, and the founders of MongoDB & KAYAK on a Software Engineer role. They're building AI-powered financial pipelines for healthcare practices, paying up to $275K + significant equity, based in NYC (onsite). Given your Python expertise, computational geometry background, and CI/CD experience shown in your repos, you'd be solving similar optimization and pipeline challenges at scale for healthcare financial data.

lol nothing says “practical applications” like a sandwich algorithm

confingy is also (finally!) the answer to the problems proposed in my old blog post bsky.app/profile/etha...

confingy is an "implicit configuration library" for Python. We're using it at work for flexibly configuring training for ML models, defining data pipelines, and more. See the post and docs for more info!

I'm happy to announce confingy, a library that I've been wanting since the start of my tech career.

You can pip install it today! (or, better yet, uv add)

Blog: www.ethanrosenthal.com/2026/03/10/c...
GitHub: github.com/runwayml/con...
Docs: runwayml.github.io/confingy/

screenshot of 3 git commits with claude + my pic on the 1st and 3rd and just me on the 2nd that says 'ruff' as the message

see, i'm still useful

Cat in the hat hats from the 90s

This might be next 😬

I’ve been seeing Jamiroquai hats on the street. I don’t know what it means, but it can’t be good.