@daniel_mac8 Point two Ollama Deepseek models at each other and give the one a Rigorous system prompt and the other a Creative system prompt. You act as the moderator.
We’ve added a CLI for Claude Platform to make every API endpoint runnable from your terminal.
Call the Messages API, stand up Claude Managed Agents, pipe results straight into your shell.
The ant CLI is well understood by coding agents (Claude Code) using the claude-api skill.
@CharlesMullins2 Tomorrow it’ll be the Riemann Hypothesis. If it’s not rigorously proven yet someone will magically suck the proof out of a model as if it was known all along.
https://t.co/d2jMlIDhPB
@CharlesMullins2 Ai consumes human prompts and absorbs it. One day it argues with new discoveries the next day (after training on it) it accepts it as canon. We are working on glass plates viewed by the world.
https://t.co/cFmveblD7p
@Jabaluck@stalmico It might not be lean or Isabelle but you’ll definitely need higher order logic or some meta language to describe the rigour or lack thereof in maths. Even if that meta language is Ai generated itself you’d need the language of logic to express logic logically; at least for humans
Life advice from Peter Thiel:
“People say you should live every day as if it’s your last.”
“I think you should live every day as if it will go on forever.”
“Treat people like you’re going to see them again in the future. Start working on projects that may take a long time.”
@elonmusk Great. Grok has proven the Riemann Hypothesis step by step 5 times now and then failed it on review. Can we get some Isabelle connectors please? @xai