AI Mondays - More Agents Is All You Need

🎙️Show Link: https://twitter.com/i/spaces/1mrxmyzAdPWxy?s=20

Who are we?

Ray intro worked at Apple for 12+ years and went from working at the Genius Bar to being in the studio with Jony Ives and being in the office with Tim Cook before presenting the Apple Watch.

Ray Fernando (@RayFernando1337) on X

Apple Engineer for 12 yrs, now at the helm of https://t.co/FvK83sCGkh. I'm learning AI in public and building apps.

https://twitter.com/RayFernando1337

Ronald Mannak

Ronald Mannak (@ronaldmannak) on X

Creator of @picoGPT. Available for iPhone, iPad, Mac and HomePod.

https://twitter.com/ronaldmannak

Trend for the week - More Agents Is All You Need

What if the solution to complex AI problems isn't more data or better algorithms, but more agents working together?

Incorporating unconventional approaches and the power of prompt engineering, we present a fresh perspective on agent-based AI frameworks.

Our analysis is grounded in practical applications, showcasing how startups and tech giants alike are experimenting with multi-agent systems.

Rapid Fire

Prompt to UI

https://x.com/galileo_ai/status/1754907661592887559?s=46&t=CSJUc-EX4J03ow_Gi8B9Lw

https://x.com/sdw/status/1755108968953954775?s=46&t=CSJUc-EX4J03ow_Gi8B9Lw

OpenAI API becomes standard: https://x.com/_philschmid/status/1755592500511997980?s=61&t=y4YUDk5rvOPclJr92oH33A

MLX 0.2.0: 40% performance gains: https://x.com/ivanfioravanti/status/1756950532457947187?s=46&t=CSJUc-EX4J03ow_Gi8B9Lw

RLX: MLX reinforced learning: https://x.com/drjimfan/status/1757065539330121955?s=46&t=CSJUc-EX4J03ow_Gi8B9Lw

$7 trillion: https://x.com/0xgaut/status/1755950984574976358?s=20

Llama lounge agents: https://x.com/jowyang/status/1745247866413830609?s=20

CrewAI

Cutting-edge framework for orchestrating role-playing, autonomous AI agents. By fostering collaborative intelligence, CrewAI empowers agents to work together seamlessly, tackling complex tasks.

https://github.com/joaomdmoura/crewAI-examples

Screenplay Writer - https://github.com/joaomdmoura/crewAI-examples/blob/main/screenplay_writer/screenplay_writer.py

Prep for a meeting - https://github.com/joaomdmoura/crewAI-examples/blob/main/prep-for-a-meeting/main.py

Paper

More Agents Is All You Need

We find that, simply via a sampling-and-voting method, the performance of large language models (LLMs) scales with the number of agents instantiated. Also, this method is orthogonal to existing...

https://arxiv.org/abs/2402.05120