Love this video about the origin of "gerrymandering": https://t.co/gXmbjQ7fWz (Gerry was a governor with a salamander-shaped district, and even he didn't like the idea) (Yes, I'm still voting for prop 50)

Going to vote yes on Prop 50 since we gotta fight stupid with stupid. Also, I have some doubts as to whether we've setup democracy correctly in this country.

Our Python+AI series is over, but you can still watch the videos, download the slides, and try out the code. Get it all from https://t.co/qbhw0u9Lir

I love when OSS projects make it *super easy* to copy the relevant version information for issue reports, like `fastmcp version --copy`.

ICE is apparently in the east bay now. Our school district has clarified it is NOT cooperating but obviously everyone is scared. They better not touch our students. (We're a sanctuary city)

🔴 Live streaming in 15 min: "MCP + Python" https://t.co/SppthLnTxi Slides @ https://t.co/x3PFkLi3N2

At #PytorchCon boothing it up, come say hi!

🔴 Live stream at 10AM PT: "Python + AI: Agents" (aka tool calling in a loop) https://t.co/ms6wLYZwUC

We just got woken up by the police knocking on our door at 5:45 AM. Apparently the cop caught college students stealing halloween decorators + wanted to see if any of the recovered skeletons were from our yard. 1) Nope, our skeletons are sad and small 2) It's 5:45 AM! WHO CARES?

🔴 Streaming at 1pm PT: Running gpt-oss via NIM on Azure serverless GPUs https://t.co/RYpSE0uLDo

🔴 Live streaming at 10 AM: "Python + AI: Tool Calling" https://t.co/xnwC6HVq1R Including tool calling in a loop, and improving tool calling reliability for SLMs. Slides:

Update on cricket-in-bed situation: confessed to partner, his only concern is being woke up by cricket. So I've concocted oat-filled cricket traps (which may actually be luxury cricket hotels) and placed on his side of the bed (given he's so unconcerned...)

Guido demo'd the approach on a DB based off his own gmail inbox, with questions about people and events mentioned in emails. My impression is that this approach could work particularly well for conversation retrieval, perhaps in combo with hybrid search.

At #PyBay25, @gvanrossum demo'd a Python package for "structured RAG". During ingestion, it uses LLM to extract structured data (entities/topics/verbs) and stores in standard DB, and then retrieves by structuring the user query as well. Try it out at: https://t.co/63YjZmKd4v

Debating how to confess to my partner that I just tried to catch a cricket, failed, and the cricket is now somewhere in our bed...

Thanks to @simonw, I can give a concrete definition for AI agents for this week's presentation about building agents in Python. No more debating, people, this is the 2025 definition! https://t.co/JqDjSKsSsu

While prepping those demos, I discovered a discrepancy with the gpt-oss Responses API vLLM implementation that causes downstream bugs with both agent-framework and langchain. We patched in agent-framework already, and a langchain PR is in progress. Good times!

I live streamed with NVIDIA today, showing how to deploy gpt-oss NIM image to an Azure serverless GPU: https://t.co/X0RROzbCT0 I showed these demos of building agents with gpt-oss and the responses API: https://t.co/mcWUs1Ee8A

It's final week of the Python+AI livestream series! Tool calling ➡ Agents ➡ MCP https://t.co/x6Cz43SJN0 We'll have code samples for every session that you can run for free, and we'll share all slides here: https://t.co/qbhw0u9Lir

@haxor is the author of "Effective Python", the book I recommend to developers that want to write more Pythonic Python and really grasp how the new Python 3 features work. It's the only Python book in our household! Get yourself a copy @ https://t.co/GOqKa2ISuy