


Prefer common sense over hype. Employed @marimo_io, built calmcode.io blog koaning.io.


The best part, clankers are really good at using this widget, so less LLM output reaches my eyelids! Made a small video to help explain what is in here. youtu.be/K5hoMJC1DeY

This widget slides over your Python function, grabs all the locals and passes that to a view function so you can see the state (of mind) of the program. It's solid for debugging, and oh boy, it is better at explaining what is going on than any clanker.

youtu.be/sOw40hTAowI Just sayin' @fastinoAI models are pretty great as a starting point to finetune decision models. Encoder-only tricks for the win!

Congrats to the team and I'm genuinely curious what they'll do next. Conductor is/was a super rad tool to work with.
I am very excited to share that Conductor is joining Meta! We have some big, crazy plans for the next interface. If you liked Conductor, you’re gonna love what’s coming. We’re super excited to work with @natfriedman and @alexandr_wang at MSL to build it. They’ve been on an absolute tear, and I’m convinced MSL has the highest talent density of all the labs. A huge thank you to our earliest believers @aaron_epstein, @ilyasu, and @nabeel. And to everyone here who tried Conductor, sent us feedback, and prompt requests - you made this so much fun. Thank you ❤️🚂 Choo choo!

I am very excited to share that Conductor is joining Meta! We have some big, crazy plans for the next interface. If you liked Conductor, you’re gonna love what’s coming. We’re super excited to work with @natfriedman and @alexandr_wang at MSL to build it. They’ve been on an absolute tear, and I’m convinced MSL has the highest talent density of all the labs. A huge thank you to our earliest believers @aaron_epstein, @ilyasu, and @nabeel. And to everyone here who tried Conductor, sent us feedback, and prompt requests - you made this so much fun. Thank you ❤️🚂 Choo choo!


ArXiv paper in -> interactive, reproducible marimo notebook out! We're thrilled to be partnering with @askalphaxiv to bring papers to life; all runnable and editable, not just static figures
GPT 5.6 Sol can one-shot convert an arXiv paper into an interactive Marimo notebook! Great for papers best understood hands on (lots of fun examples in interpretability, inference engineering, agent harnesses, benchmarking, and more) Play around with the notebook, inspect the code, or try the same workflow with your own agents below

I gave a demo yesterday to my colleagues that went so well that I figured I might as well make a recording of it right away. It's super early, but it's a v0 in a direction of "more humane coding experiences". May it offer you something to think about! youtu.be/BrQLCqofc30

marimo webapps can also be deployed as interactive slides. And that means that if somebody asks a question about the data ... that you can live update the SQL behind the chart! Check it out if you haven't already: youtube.com/watch?v=_jcR8sje…

Sometimes you win by throwing a GPU at a problem. But even in those cases ... it helps to consider your throwing technique. More details, and fun algorithm, in this video: youtu.be/1-5wkzdP1Jc


72 hours left to win a @FrameworkPuter Laptop or Mac Mini + @claudeai subscriptions. Not too late to start: last round, almost 100 entries came in on the final day. Pick any @askalphaxiv paper, build in molab, & submit by July 9, 11:59 PM PT 👀 marimo.io/pages/events/noteb…

Made with #python #numpy #matplotlib and #marimo (and also #wigglystuff)

TIL about elasticnet. An energy function trick to attempt to make TSP a differentiable problem. It doesn't perform well, but the animation is pretty! molab.marimo.io/github/koani…

This video is all about a silly experiment that's been in my mind recently. Now that molab has GPUs ... can I use deep learning architectures to solve (NP)-hard computer science problems? youtu.be/QnP-EdJyiDo?si=6sqx…