What's (actually) up with o1
The new o1 UX is bad. The model is weird. I have questions.
OpenAI
The GPT store is stupid and dead
There’s no moat. All prompts can be extracted, so all prompts are public.
How Sora (actually) works
ChatGPT can now remember custom instructions across prompts. Here's how to use it.
You can now make ChatGPT follow custom instructions for all conversations.
New study validates user rumors of degraded GPT-4 performance
Monitoring the Drift in Large Language Models: An Examination of GPT-3.5 and GPT-4
A Plain English Guide to Reverse-Engineering Reddit's Source Code with LangChain, Activeloop, and GPT-4
Reverse-engineering Reddit's public source code using GPT-4 and Activeloop with Langchain code understanding
Getting Started with the Vercel AI SDK: Building Powerful AI Apps
Vercel is launching new tools to improve how you work with AI