AI Devs love Python – 50% of 300K public AI projects on Replit
– 32M forks of Replit’s main Python template
– 3.7M forks of the Python with Turtle template
– 168K Devs taking 100 Days of Python https://
x.com/Replit/status/
1679898663831879680
…
INNOVATION
-

Python Dominates AI Development: 300K Projects on Replit
By
–
-
Self-play and reinforcement learning in current AI systems
By
–
With current AIs, you do need to set that up to happen and it's not on by default. That said, "self-play" is how AlphaGo, Alpha Zero, etc. learned to play Go; more generally the entire field of "reinforcement learning" is this.
-

Paul Graham Interview with Tyler Cowen on AI and Innovation
By
–
Such a fun and delightful interview with @paulg by master interviewer @tylercowen https://
open.spotify.com/episode/6QQr6e
O9qgQTwdo7UbdiXq
… -

Evaluating Question-Answering Applications: A Practical Cookbook
By
–
Everyone's building question/answering applications, but evaluating them is pretty tricky We're trying to make this easier. This cookbook shows how to evaluate the final answers (end-to-end app). We'll add one more focused on just retrieval shortly. What else would be helpful?
-

Lennybot reaches 150 daily users with quality questions
By
–
Seeing about 150 people using http://
Lennybot.com each day, and seeing some really good questions! How fun! -
Cloud Companies Improving Open Source Support and Licensing
By
–
Years later, some sw companies have gotten better at operating it and AWS (among other clouds) has gotten better at contributing to the sw. But there's still a gap and that prompts things like the BSL.
-
Cloud Giants’ Advantage in Monetizing Open Source Software
By
–
Years ago @timbray called out the divide between the cloud vendors (awesome at operating sw at scale) and others better at building sw. https://
tbray.org/ongoing/When/2
01x/2019/09/21/AWS-Open-Source-People
… That divide is challenging to navigate bc it's easy to feel ill-used by, eg, AWS for being better at turning oss into $$ -
HashiCorp’s Shift: Open Source Commitment Under Scrutiny
By
–
Et tu, Hashi? It's always been overly façile to blame "VC-backed biz", and claim companies just didn't get open source/lacked commitment. But Hashi was (and is!) one of the "good guys." So this isn't done out of some "We're not rich enough" motive. Time to confront reality https://
t.co/VlrubyAK9P -

LangSmith Enables LLM App Debugging Without LangChain
By
–
Woah! With LangSmith, you can now debug, test, evaluate, and monitor your #LLM apps, even if you aren't using @langchain
! Great new guide with @streamlit baked in!