A coworker on GPT-4's mistakes with detailed list of instructions: "it makes the same variety of mistakes that a team of 1,000 different humans would make". That is, it covers the full distribution of reasonable interpretations rather than picking one — a strength and a weakness!
AGI
-
Awesome Generalist Investor Approach to AI
By
–
I like to think of myself as an Awesome Generalist Investor
-
Multiple Autonomous Agents Working Together Demonstrated
By
–
Apr 6
And we're now seeing multiple autonomous agents work together… -
Baby AGI term emerges amid market skepticism and debate
By
–
Mar 31
Doomers came out in full force over the next few days. It's where the name "baby AGI" comes from. -
Building an AI Founder Prototype: Exploring Autonomous Innovation
By
–
Mar 26th
— Yohei (@yoheinakajima) 7 avril 2023
Two Sundays ago, I was playing around w the idea of building an AI founder and shared this prototype on Twitter… https://t.co/rgKIaMCLz6Mar 26th
Two Sundays ago, I was playing around w the idea of building an AI founder and shared this prototype on Twitter… -
AGI Jokingly Asked to Make Paperclips Goes Viral
By
–
People started calling it AGI, so I jokingly asked it to make paperclips which got picked up by the man himself…
-
GPT-4 Updates Priors on AGI Development Progress
By
–
We don’t have AGI yet but GPT4 has definitely caused me to update my priors. So consider that piece obsolete while we all learn more.
-
Building Self-Learning Web Search Agents with RLHF Fine-Tuning
By
–
Start w few tools to figure out what works/doesn’t. Use that knowledge to build a web search agent that can figure it out on its own. Use this to build an auto-updating db of all possible tools w documentation. Then fine tune a model w RLHF to decide which tool to use.
