ChatGPT o1-preview explains that LLMs cannot reliably count or spell, and are not appropriate tools for such tasks, in a poem where the first letter of each word spells: "Scalable inference is all you need; Strawberry fields forever."

I’m so jealous

Yayyy new BdoubleO100’s video

hold on. don’t grow tired. don’t give up. HE is better.

cyber truck monster truck

Have a question that is challenging for humans and AI? We (Center for AI Safety + Scale AI) are launching Humanity's Last Exam, a massive collaboration to create the world's toughest AI benchmark. Submit a hard question and become a co-author. Best questions get part of $500,000 in

qUaDrILliOn

many to one relationships with ai. this is awesome. I want to build this but make it a twitch chat

gonna cook today

this is pretty insane. Maybe o1 was a large step.

someone should have sonnet 3.5 "think about" a problem/task for 100 million tokens and see if it cooks up something insane. (probably a loop of "consider all angles", "suggest novel outside thoughts and counterfactuals", summarize, go to step 1) cc: ampdot j⧉nus

this is genius use of notebooklm

I tested Pixtral, the new vision model just released by Mistral AI It's surprisingly good! The best OSS vision model I've tested so far. Check out the full test:

why isn't pixtral on ollama yet

here’s an llm usage tip that most people aren’t leveraging: when you’re using an llm to generate content that you plan to copy paste like answers to questions or ai art prompts or whatever, tell the model “put each answer in its own markdown code block” copy and pasting easier

astute observation. both is best.

“I don’t even see the R’s. All I see is 302, 1618, 19772, 198, 3504, 1134, 19772, 198, 101830, 198, 138322, 198, 1100, 302, 1618, 19772, 25644, 1100, 3504, 1134, 19772, 1100.”

Sully With each passing day my opinion strengthens that GitHub needs to add a “copy all files in this repo at once to clipboard” button

The o1 team OpenAI put out a long-form conversation about the creation process and their favorite things to use it for that other models can't do. I used o1-preview to extract that list of use cases from the video using Fabric.

🤔 What do you get when you combine Prompt Injection + Data Exfiltration + Long-term Memory? 👉 Persistent SpAIware! 🚨 OpenAI fixed a persistent data exfiltration issue in the macOS ChatGPT app that I reported. 🔒 Make sure to stay up to date and use the latest ChatGPT app