Kyle

On Thursday I was interviewed by Demetrios Brinkmann as part of the @aaif.io livestream where I got to talk about the more interesting aspects of writing a book about a rapidly changing technology, the serendipitous timing of all my blown deadlines, what the book is about, and more

On Thursday I was interviewed by Demetrios Brinkmann as part of the @aaif.io livestream where I got to talk about the more interesting aspects of writing a book about a rapidly changing technology, the serendipitous timing of all my blown deadlines, what the book is about, and more

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

“What game ought we be playing in AI?” We are being sold a winner takes all game:

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

The importance of the OSS old heads to building the ideological foundation for the next generation of builders to carry forward and ensure technology remains in the hands of people as something that can empower and enrich us all (and not just economically) can’t be understated.

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

This formed an “architecture of participation” - a small kernel surrounded by cooperating utilities. Anyone can add or modify these parts using a set of rules - a protocol-centric OS. He posits that MCP brings this AoP to AI, pulling against the winner-takes-most design by the frontier AI companies

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

Want to build for Claude products? Note the open core:

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

Now Kelly Fang on how @anthropic.com is using MCP inside the org. They have built a Claude directory to show all of the connectors and plugins that work across all Claude tools. Unclear if this is related the MCP server directory or a sister tool.

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

Governance (identity, registry, policy, and observability) for all agents inside of an enterprise.

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

It’s important to find your own universal properties:

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

Next up is Dr. Shazam Laddad from AWS on how to verify code correctness. He introduces “universal properties” - specifications that apply to *any* application regardless of implementation. Strongly typed languages have this, eg rust preventing multiple borrows.

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

He does point out that we need a wider traceable unit, but I missed that slide. Still, our infrastructure must be decentralized. We are using a centralized tool for decentralized git, we’ve largely lost the advantages of decentralization.

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

accepted without human review. This is bad! But…the SDLC as we know it is dead. It was so focused on making loops intelligible to humans while driving down the size of those loops. But agents are now driving these. So what do we do to make sure our intent travels on the loops with agents?

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

The real thrust of this talk is how to work effectively across disciplines. Emphasizes that the human is the one that has to “be the agent” to make it happen.

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

A lot of this work is, unlike the popular coverage, about what you’d expect for personalized vaccine development. LLMs seem to have largely been used as personal learning assistants. A very good intro to how modern drug discovery works.

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

Already, he is speaking of building ontologies and doing things we’ve done in healthtech and biotech for doing this work - a good thing not covered in any of the popular coverage of what he and his team did. An early idea was “finding the diff” between baseline DNA and cancer cells. Not a new idea,

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

“Coordination could change the scale at which people organize work”

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

A theme I’ve seen in a few talks about agent-agent communication is that agents can carry requests from its owners but never their authority. Wang proposes a broker in between agents to do this.

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

The PyTorch Foundation’s goal is to build models on any chip, any cloud, with any agent. Now Chi Wang formerly of Google DeepMind and now at AG2 to speak about how we can get personal agents to work together.

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

Jonathan shares some history of model building focus, starting with pretraining, then post-training, and now “on the job” training - retraining models regularly with feedback from agents. This isn’t just about LLMs, but this shows the importance of continual learning. Right now the loops aren’t

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)

And the keystone award winner is…🥁 Den Delimarsky of @anthropic.com and lead maintainer of MCP

Going to attempt to do more chill coverage right now of the AGNTCon + MCPCon morning keynotes because I’m sleepy, but because it’s dark in here these threads are also my notes for myself (I’m back to doing hand-written notes because LLMs can read my handwriting better than I can)