What are the key takeaways from this Practical AI episode?

Key insights include: **Agent Skill Accumulation:** Hermes Agent automatically creates reusable skills without explicit programming. When the agent solves a novel problem — such as navigating bot-detection systems to book a restaurant reservation in 45 minutes — it logs the method as a named skill. Subsequent identical tasks complete instantly by retrieving that stored skill, compounding efficiency the longer the agent runs.; **Outcome-Oriented Prompting:** Users should describe desired end states and explicit evaluation criteria rather than step-by-step instructions. Unstated assumptions about quality — tone, format, aesthetics — will never be fulfilled because the model defaults to statistically average outputs. Writing out every condition you would use to judge success gives the agent a concrete target to optimize toward.; **Agent Deployment Targeting:** Identify workflows requiring infinite patience but minimal creativity as the highest-value automation targets. Tasks a non-specialist human could theoretically complete but would never willingly sustain — such as reading thousands of server logs to root-cause errors — are ideal. Avoid one-to-one role mapping like "CEO agent," focusing instead on specific operational processes.

What did Hermes Agent discuss on Practical AI?

Nous Research cofounder Jeffrey Cannell explains how Hermes Agent, now the number one open source repository on GitHub, was built internally as a model research tool before being released publicly, and how its self-improving skill and memory systems differentiate it from competing agent harnesses in the market. Key topics include: **Agent Skill Accumulation:** Hermes Agent automatically creates reusable skills without explicit programming. When the agent solves a novel problem — such as navigating bot-detection systems to book a restaurant reservation in 45 minutes — it logs the method as a named skill. Subsequent identical tasks complete instantly by retrieving that stored skill, compounding efficiency the longer the agent runs.; **Outcome-Oriented Prompting:** Users should describe desired end states and explicit evaluation criteria rather than step-by-step instructions. Unstated assumptions about quality — tone, format, aesthetics — will never be fulfilled because the model defaults to statistically average outputs. Writing out every condition you would use to judge success gives the agent a concrete target to optimize toward..

How long is this episode of Practical AI?

This episode is 51 minutes long. SignalCast provides an AI-generated summary so you can get the key insights in about 3 minutes.

Practical AI

Hermes Agent: Agents that grow with you

May 21, 2026

51 min episode · 2 min read

Hermes Agent

Episode

51 min

Read time

2 min

Topics

Productivity, Investing, Startups

AI-Generated Summary

Published May 21, 2026

Key Takeaways

✓Agent Skill Accumulation: Hermes Agent automatically creates reusable skills without explicit programming. When the agent solves a novel problem — such as navigating bot-detection systems to book a restaurant reservation in 45 minutes — it logs the method as a named skill. Subsequent identical tasks complete instantly by retrieving that stored skill, compounding efficiency the longer the agent runs.
✓Outcome-Oriented Prompting: Users should describe desired end states and explicit evaluation criteria rather than step-by-step instructions. Unstated assumptions about quality — tone, format, aesthetics — will never be fulfilled because the model defaults to statistically average outputs. Writing out every condition you would use to judge success gives the agent a concrete target to optimize toward.
✓Agent Deployment Targeting: Identify workflows requiring infinite patience but minimal creativity as the highest-value automation targets. Tasks a non-specialist human could theoretically complete but would never willingly sustain — such as reading thousands of server logs to root-cause errors — are ideal. Avoid one-to-one role mapping like "CEO agent," focusing instead on specific operational processes.
✓Organizational Skill Compounding: A single engineer using Hermes Agent to debug infrastructure problems passively builds a shared organizational knowledge base. After one month of use, Nous Research's support staff could independently query complex backend systems without engineering involvement, because the agent's accumulated skills became accessible across the entire team without additional training or coding.
✓Western Open Source Shift: NVIDIA's commitment of approximately 20 billion dollars toward training Western open source models over several years represents a structural change in who funds open AI development. Because all AI workloads ultimately run on NVIDIA hardware, the company has direct business alignment with expanding open source model usage — unlike Meta, whose Llama program lacked equivalent commercial incentive.

What It Covers

Nous Research cofounder Jeffrey Cannell explains how Hermes Agent, now the number one open source repository on GitHub, was built internally as a model research tool before being released publicly, and how its self-improving skill and memory systems differentiate it from competing agent harnesses in the market.

Key Questions Answered

•Agent Skill Accumulation: Hermes Agent automatically creates reusable skills without explicit programming. When the agent solves a novel problem — such as navigating bot-detection systems to book a restaurant reservation in 45 minutes — it logs the method as a named skill. Subsequent identical tasks complete instantly by retrieving that stored skill, compounding efficiency the longer the agent runs.
•Outcome-Oriented Prompting: Users should describe desired end states and explicit evaluation criteria rather than step-by-step instructions. Unstated assumptions about quality — tone, format, aesthetics — will never be fulfilled because the model defaults to statistically average outputs. Writing out every condition you would use to judge success gives the agent a concrete target to optimize toward.
•Agent Deployment Targeting: Identify workflows requiring infinite patience but minimal creativity as the highest-value automation targets. Tasks a non-specialist human could theoretically complete but would never willingly sustain — such as reading thousands of server logs to root-cause errors — are ideal. Avoid one-to-one role mapping like "CEO agent," focusing instead on specific operational processes.
•Organizational Skill Compounding: A single engineer using Hermes Agent to debug infrastructure problems passively builds a shared organizational knowledge base. After one month of use, Nous Research's support staff could independently query complex backend systems without engineering involvement, because the agent's accumulated skills became accessible across the entire team without additional training or coding.
•Western Open Source Shift: NVIDIA's commitment of approximately 20 billion dollars toward training Western open source models over several years represents a structural change in who funds open AI development. Because all AI workloads ultimately run on NVIDIA hardware, the company has direct business alignment with expanding open source model usage — unlike Meta, whose Llama program lacked equivalent commercial incentive.

Notable Moment

Cannell revealed that Hermes Agent was originally built by a Nous Research team member with minimal coding experience, using AI tools to architect the entire application. That project became the top open source repository on GitHub, illustrating how AI now enables non-developers to build production-grade software infrastructure.

Know someone who'd find this useful?

Episode Transcript

Welcome to the Practical AI podcast, where we break down the real world applications of artificial intelligence and how it's shaping the way we live, work, and create. Our goal is to help make AI technology practical, productive, and accessible to everyone. Whether you're a developer, business leader, or just curious about the tech behind the buzz, you're in the right place. Be sure to connect with us on LinkedIn, x, or Blue Sky to stay up to date with episode drops, behind the scenes content, and AI insights. You can learn more at practicalai.fm. Now onto the show. Welcome to another episode of the Practical AI podcast. This is Daniel Whitenak. I am CEO at Prediction Guard, and I'm joined as always by my co host, Chris Benson, who is a principal AI and autonomy research engineer. How you how you doing, Chris? Hey. Doing great today. How's it going, Dane? It's it's going really good. I've I've been excited for this one, I guess, ever since a conversation that we had back in 2024 on the show with one of the cofounders of News Research. Today, we've got another one of those cofounders with us, Jeffrey Cannell, who is cofounder and CTO at News Research. Welcome. Hi there. Thanks for having me on. Yeah. It's it's great to great to catch up. Of of course, big admirers of of a lot of the things you all are doing and and really enjoyed the last conversation. That was, though, in 2024. So it's been a couple years, and that's, like, eternity in the Twenty years in AI time. Yeah. Yeah. Yeah. Twenty years in AI time. So I I believe, if I remember right, kind of at that time, news research, which for for listeners, I would encourage you to go back and listen to that episode, which was, at at a certain point, primarily a kind of distributed, community of folks on Discord and and elsewhere, but I believe that was kind of morphing into more of a actual company or or organization at at that time. Could you kind of help us understand how that's evolved over the last couple years, maybe some highlights for from the news history books over the over the past couple years? Yeah. Like you said, it was two years ago, so twenty years in AI time. And we did sort of start out, like you said, as sort of this loose collaboration of, the homies on the Discord as I always like to say. And, that sort of, grew organically just through Twitter and other places like that. People who are fundamentally interested in open source AI and and making sure that this, you know, this tech that we think is, you know, the transformational technology of our of our of our generation wasn't gonna just get locked behind a few closed companies. And at the time, we were sort of looking up at, you know, GBD 3.5 had just come …

Get the full transcript (10,532 words) + summary by email — free

One-time email with the complete transcript and AI summary of this episode. No account needed.

One email, no spam. We’ll also show you what SignalCast does.

Browse all Practical AI transcripts →

You just read a 3-minute summary of a 48-minute episode.

Get Practical AI summarized like this every Monday — plus up to 2 more podcasts, free.

Pick Your Podcasts — Free

Books, tools, and gear mentioned in this episode

SignalCast may earn commission on purchases via these links.

Tools

Hermes AgentRecommended
by Nous Research
“Nous Research cofounder Jeffrey Cannell explains how Hermes Agent, now the number one open source repository on GitHub, was built internally as a model research tool before being released publicly, and how its self-improving skill and memory systems differentiate it from competing agent harnesses in the market.”
Prediction Guard
by Prediction Guard
“SPONSORS: Prediction Guard (https://predictionguard.com/practicalai)”
Framer
by Framer
“SPONSORS: Framer (https://framer.com/practicalai)”