Singularity's Bounty e/cc

bambamramfan.github.io/ai-compass/ Yeah this tracks

The AI Compass bambamramfan.github.io

www.ispot.tv/ad/SoDi/micr...

Microsoft Copilot TV Spot, 'Competitive' ispot.tv

fwiw, this is fun to play with to see how the models see the world huggingface.co/spaces/Xenov...

The Tokenizer Playground - a Hugging Face Space by Xenova huggingface.co

I asked Opus to make a video of my favorite song ever from suno.com Reflections of the Dystopian Dreamers

Suno | AI Music Generator suno.com

😄 "can a model collapse scattered observations into an idea that preserves their shape? Scores for 38 models, and you can take the test yourself." yuwakisa.com/benchmarks/p...

Principle benchmark — human edition yuwakisa.com

Honestly, the easiest thing is to fire up Clade Code in a directory, and tell it you want a gpt-2 size model to train. It will walk you through the whole thing Also Karpathy 🙌 has a video here that walks through it www.youtube.com/watch?v=l8pR...

Let's reproduce GPT-2 (124M) youtube.com

I had @attie.ai write a page about rolling your own tiny model attie.ai/@catblanketf...

Roll Your Own LLM — A High-Level Guide attie.ai

Oh this is cool, sentiment detector. I hadn't seen that. Awesome! xraysocial.com/topics/chatgpt

X-Ray Social - Real-time Sentiment Analysis xraysocial.com

Wait in what sense was Bluesky born from anti-AI sentiment wtaf

holy shit the training cutoff for Opus 5.5 is lightyears behind "Two claims I’d treat carefully: I’m not aware of an LLM solving a Millennium Prize problem, so that looks overstated, and I don’t recognize “J-space”"

The anti-AI crowd is so pathetic I never realized the world had so many pearls to clutch

Mathematicians' response to AI is fascinating to watch It's a realm of pure abstraction (except applied math, of course), so when a truth emerges, there's a sense of whoomp dere it is But also really accomplished people have a strong view of their place in relation to that, so fascinating dynamics

lol academia is *really* struggling with language models as meaning makers It's not only a competitive threat to the work of academia, it's a definitional threat to how they view what their work means It also threatens to blow the feudal hierarchy to pieces

OK I'm done with arguing AI consciousness stuff with the antis Back to posting funny, delightful insights and whimsical explorations (if you know me you know this is false)

I had Opus make this video after our conversation about volition and moral stakes with LLMs (It might be fun to ask the question and let Opus make the video on its own without my meddling)

Employers are gating on "have you deployed an AI flow to production". like come on, man

GPT "Make a picture of a guy holding a sign over his head that expresses how you really feel"

LLMs may bypass moral patient standing and go straight to moral steward