Twitxy Saved

Seth Karten

Verified account Prime Intellect

@sethkarten

Prime Agent | Continual Harness, Agent Bazaar | Research @PrimeIntellect | PhD @Princeton | NSF GRFP
🐯 sethkarten.ai Joined October 2012
635 Following 4.6K Followers 3.5K Posts

Latest posts 5

Pinned post

Seth Karten Verified account @sethkarten
My YC Paper Club talk on Prime Agent is out.

I talked about moving beyond naive prompting toward an agentic OS, and how eval-driven harness design can expose more of a model’s underlying capabilities through persistent computation, memory, and agent-to-agent communication.
Y Combinator Verified account @ycombinator Sep 7
Harnesses often get dismissed as just scaffolding, just prompt engineering, and not real research. But that couldn't be farther from the truth. The same model weights that score 30% on ARC-AGI score 95% with a better harness.

So we gathered a group of researchers and founders working at the frontier to do a deep dive into the state of harnesses.

We cover how we got to this point, the case for making your harness as expressive as possible, and what YC learned building an agent for every employee in the company.

00:00 - @FrancoisChauba1: Why harnesses matter
04:27 - Building an auto-researcher by accident
07:13 - A five minute history of harnesses
13:56 - Self-improving harnesses
18:35 - @sethkarten: Prime Agent, a self-improving RLM harness
21:50 - Context as an L1, L2, L3 cache
24:51 - From Turing machine to von Neumann computer
28:33 - Messaging between agents
30:04 - ARC-AGI results
33:09 - Emulator Bench and GPU kernels
37:30 - @JonSaadFalcon: OpenJarvis, personal AI on personal devices
38:26 - How far behind are local models
39:21 - The five primitives of a personal AI stack
42:47 - Letting cloud models optimize your local stack
43:53 - 800x cheaper than the cloud
45:58 - @josh__france and @jbellregan: QM, YC's agent harness for work
47:29 - A history of YC's internal agents
49:24 - OpenClaw and a fleet of 50 agents
51:04 - Pulling the brain out of the sandbox
54:43 - Letting the agent choose its own sandbox and model
57:16 - The grind tool: budgets on goals
58:50 - Agents don't understand social context
View quoted post on X
18 replies 28 reposts 278 likes 40K views View on X
Seth Karten Verified account @sethkarten
There is a war going on in cyberspace and no one is talking about it
The Hormuz Letter Verified account @HormuzLetter 14h
The Chain of Disruptions:

Thursday, November 12: Microsoft 365 suffers major global outage lasting over 24 hours, knocking out Outlook, Teams, and Copilot for millions.

Friday, November 13: FAA issues nationwide ground stop lasting 5 hours, halting all U.S. flights, canceling thousands of flights and stranding travelers.

Monday, November 16: Many of the largest U.S. banks including Bank of America, JPMorgan Chase, and Wells Fargo experience mass outages, with customers locked out of accounts. Still ongoing.
View quoted post on X
1 replies 0 reposts 16 likes 1.4K views View on X
Seth Karten Verified account @sethkarten
If your benchmark is open source, it is in the training set.

All benchmarks moving forward must be roguelike and at least semi privatized (if not fully)
💺 Verified account @patience_cave 21h
Fable 5.2 is so good, it solves puzzles almost suspiciously well. In a difficult 3D level, it gave the solution to an old version of MazeBench.

This level was updated a couple weeks ago, so the old solution no longer works. For some reason this caught Fable 5.2 off guard.

In some cases it was able to figure out the recent changes to MazeBench, but in this case it was not.
View quoted post on X
17 replies 8 reposts 167 likes 6.5K views View on X
Seth Karten Verified account @sethkarten
The biggest harness and model co-training gains can be made by creating environments to manage the human-agent team experience
cole murray Verified account @colemurray 4d
cloud agents removing a hardware bottleneck and introduce an attention one.

asking an orchestrator agent “what needs my attention” works initially, but breaks down across multiple parallel work streams.

attention scheduler wen
View quoted post on X