If you're interested in Reinforcement Learning, stay alert - we figured out some stuff🔋!! Built by Lukas Ego, Linus Meierhöfer, Josef Zoller, and Hannes Furmans

Wrangling prompts is the most frustrating part of building AI tools right now. The user should be able to say "Yes, this was accurate", "This response wasn't good", or "I would have phrased it like this..." The AI should then be allowed to edit its own prompt or include these

Tom Blomfield Agreed! That's exactly what we're building Augento for 🤖 - alignment through RL

You can now download the models you trained with us! augento.ai/blog/download-…