Tab CLearning (September 23 2024)
Tab Clearing Sept 23
https://openai.com/global-affairs/openai-academy/: “Investing in the development of local AI talent can fuel economic growth and innovation across sectors like healthcare, agriculture, education, and finance.”. Thought: Educating people how to leverage AI is a high-impact thing to do today
https://ia.samaltman.com/ “The Intelligence Age” tl;dr sam thinks the future is going to be cool
https://team.doubao.com/en/special/seed-music - people working on audio models, I like to see it! This Chinese team have a decent-looking one with various ways to control it.
https://github.blog/news-insights/product-news/openai-o1-in-github-copilot/ - o1 is, predictably, cool for slightly more complex tasks.
https://blog.adobe.com/en/publish/2024/09/11/bringing-gen-ai-to-video-adobe-firefly-video-model-coming-soon - Firefly video is looking cool, great for B-roll I guess, or realistically pre-viz. Generative extend seems like the most useful bit. They make a big deal of commercially safe, not trained on user data vibes
https://www.hume.ai/blog/introducing-evi2 Guys bossing around vitual assistants for demos, making them sing etc feels icky. But voice is interesting to me, I wish OpenAI’s voice mode was available. I might try hume evi2, will see how it does.
https://rachel.fast.ai/posts/2024-09-10-gaps-risks-science/ Always appreciate Rachel’s posts
https://blog.mariusvach.com/posts/passwordless-auth-fasthtml So glad people are writing up things like this, the FastHTML ecosystem is growing
https://kingbird.myphotos.cc/packing/squares_in_squares.html - I want to know how big to make the box for printing the 17-square packing problem.s=4.67553…
https://gist.github.com/simonw/d8cc934ad76b3bba82127937d45dc719 - The System prompt for val.town/townie - a good reference for any similar systems.
https://github.com/btsmart/splatt3r - directly predicts 3D gaussians from a couple of uncalibrated images, very cool
https://github.com/hudgeon/fasthtml_supabase_auth_example - glad it exists, haven’t looked yet
https://transformer-circuits.pub/2024/august-update/index.html - glad someone is tinkering with this stuff
https://blog.eleuther.ai/mutransfer/ - This post is pretty good. It goes into the details of mutransfer as the title suggests. They have a really nice nano GPT implementation that shows the changes in action. I think this is something that should hopefully be picked up by everyone going forward.
https://osf.io/preprints/osf/mtgn7 - “Why Is Anything Conscious?” - Fun to skim, I liked the ’From Rocks To Einstein” section but skipped a lot of the formalisms and silliness. 