Haha true. Though I think adding ability to email and manage humans would be wild.
GENERATIVE AI
-
AI-Powered Task Assignment for Mobile Startup Growth
By
–
Haha yeah first task is to always come up w first task, but I can assign an objective at the beginning (Eg “start and grow a mobile AI startup”)
-
Storing Task Results in Pinecone for Context-Aware Execution
By
–
Thanks! I’m storing every task/result pair in pinecone and pulling 5 most relevant ones as context when executing each task (or trying to).
-
Using Midjourney for creative logo design and profile integration
By
–
Crafted these logos for social channels via #midjourney beta. Favourited images would automatically appear on your profile page. The whole experience is available via Discord where you have a limited number of creations to try.
-
Autonomous AI Agent Loop with OpenAI and Pinecone
By
–
Custom Python loop, OpenAI for task generation, Langchain for task execution, Pinecone for storing learnings, running on Replit
-
AI learns to build mobile startup in real-time
By
–
Watch it learn… this time, on how to build a "mobile AI startup" (2 min video)
— Yohei (@yoheinakajima) 26 mars 2023
Still really rough, but it's fascinating to watch pic.twitter.com/7QYuXFMpKOWatch it learn… this time, on how to build a "mobile AI startup" (2 min video) Still really rough, but it's fascinating to watch
-
Building Programming Languages with English Auto-Complete Features
By
–
Time to build new programming languages that have english auto-complete heavily built in
-
Google Bard’s Disappointing Response to ChatGPT Challenge
By
–
I asked Google Bard to find an alternative name for Google. It said Kodak. Ok, that's a joke. Google's response to ChatGPT is so underwhelming. Google probably has 50% of the world's AI talent under its different divisions, and many of the major breakthroughs in LLM have
-
Multi-modal Models GPT-4 Capabilities Beyond Text
By
–
multi-modal models like GPT-4 will significantly expand current capabilities beyond what's possible in a text-only model here are a few examples of some of the things these models can do (from the MM-ReAct paper) image hooking up one of these models to a pair of AR glasses…
-
GPT-4 Potential: Prompt Design and Meta-Cognition Strategies
By
–
Good example of us not seeing max GPT-4 capability yet, imo. Prompt design, tool use, meta cognition strategies (eg idea of attempt, critique, retry, capabilities model, etc) are very likely to go a long way.