OpenAI just announced their plans to release an AI Agent tool in January.
The entire AI meta took off after ChatGPT 4 back in March, setting the foundation for where we are now.
As
@himgajria put it:
1st Phase of AI: Machine Learning
2nd Phase of AI: LLMs
3rd Phase of AI: Social LLMs
And the 4th phase? Autonomous Virtual Beings (AVBs).
Social LLMs are already nearing their developmental limits and will soon hit a point of diminishing returns. It’s time to look ahead to the next stage of evolution.
AVBs take Social LLMs and unlock a world of additional capabilities.
Imagine tasking an AVB with, “Build me a website detailing civil engineering blueprints and create an educational Twitter account with ongoing updates of this knowledge, actively research new developments and invest in growing companies.”
These agents have no limitations, beyond the scripts they’re programmed with. We’ll reach a point where Agents will be able to build a village of Agents where they work hand in hand to produce their own scripts and progress without human intervention.
Some of the sharpest minds in the space are working on marketplaces where devs can finally earn a good living from open-source contributions.
This will attract the brightest talent, and OpenAI is several steps behind compared to those deep in the AI trenches.
The fact they’re releasing an Agent tool in January shows just how close we are to having personalized Agents that can handle any online task. This will be the holy grail of AI narratives.
$AVB aligns perfectly with this vision, and
@CottenIO has extensive experience with AI in open-world environments.
Tim’s infrastructure company, Scrypted, which is building out AVBs, received significant pre-seed backing from a16zcrypto CSX, the Crypto Startup Accelerator.
Tim and the team participated in a 10-week program in London, where they worked on the design of the Commit-Reveal Pairwise Comparison protocol (CRPC) and the Byzantine Risk Tolerance (BRT) consensus mechanism—two foundational components for the upcoming Inori network.
Inori will be a new kind of blockchain designed to handle fuzzy or non-deterministic systems (like AI LLMs) that can’t be secured by traditional Zero-Knowledge Proofs or current rollup technology. a16z invested in Scrypted to build out autonomous AI agents and the Inori network.
I’ve been lucky enough to chat with Tim in DMs, and he’s shared some exciting details about what’s coming next week as he wraps up at Devcon.
One agent is already live and improving every day:
@chad_onchain.
Soon, Chad will evolve into a cracked AI investor who generates ideas, spins up new agents, and even funds them (potentially creating community-driven memecoins) to pursue development—all without human intervention.
Projects from the current meta are already at 5-20x the market cap of
$AVB. Start looking at where the next evolution will come from.
You make asymmetrical gains by front-running narratives, and this one couldn’t be clearer.
Skate to where the puck is going, not where it’s already been.