The repeated blind spot I see with people about AI is that their main reply is "it's not good enough" and "not passable as human" etc. But that all implies millions of people aren't working to improve it every day If you wait until it's good enough you're too late

Autocomplete for #AIart #prompts! I built Teleprompter, a Recurrent Neural Network trained on 50,000 of the best #Midjourney prompts. I linked it with #GPT3 to further refine the outputs. Link to the LIVE app and examples in comments 👇 RT for more AI projects! #ai

Třetí místo jde GPT robotických pomocníkům, teda týmu, který je vytváři.

Do you know Arc - Arc browser from The Browser Company? It has a cool feature "Browse for Me". I used Perplexity but must say that the "Browse for Me" feature is divine and in a different league than Perplexity, which didn't handle my requests well.

gemini 1.5-pro-exp outperforms gpt-4o. Some say it's the end of OpenAI reign. So I ran my own tests: test #1 → write a viral twitter thread

New model on arena again. It is simply named 'anonymous-chatbot' this time. There is speculation already that this is Q*. Whatever it is, something new is on the way.

🆕 Prompt caching with Claude. Caching lets you instantly fine-tune model responses with longer and more instructive prompts—all while reducing costs by up to 90%. Available in beta on the Anthropic API today. anthropic.com/news/prompt-ca…

Today we introduced Claude for Enterprise - the best way to securely work with Claude within your org. We're also launching a beta GitHub integration for early Enterprise users. We plan to make this more broadly available to Claude dot ai users later this year.