Nate Herk | AI Automation
Hi 👋 I'm Nate Herk. I’m passionate about helping businesses unlock the power of AI Automations to save time, cut costs, and stay competitive. I believe that businesses that don’t adopt AI will struggle to keep up, and I’m here to make sure you’re ahead of the curve. Whether you’re new to AI or looking to refine your workflows, I’m here to guide you through it all. Let’s build a more efficient future together! Book a call on my website!
Finally. Agent Loops Clearly Explained.
The video explains what agent loops are and why they matter: a loop is a recursive process where a system reasons about a goal, acts, observes results, and repeats until a defined done condition is met. It emphasizes two core pillars—clear objective (done criteria) and robust verification—and shows practical examples from loop libraries and simple solo loops, highlighting when and how to use loops, the importance of objective scoring, and the trade-offs between continuous automation and human-in-the-loop checks.
GLM 5.2 in Claude Code is Blowing My Mind
The video compares GLM 5.2 (Cloud Code) to Opus 4.8 through hands-on demos, showing faster setup, lower cost, and solid performance on a variety of tasks. It covers practical setup steps (billing, API keys, environment config), real-world examples (storm research, HTML reports), and guidance on when to use open‑source models versus closed‑source options based on task requirements and cost.
1000+ Hours in Claude Code in 60 Mins
The chat centers on turning cloud code into a powerful, AI-native system by building an orchestrated, agent-based workflow (the harness) rather than just using AI tools in isolation. It covers planning, verification, and security, emphasizing four pillars (planning with context, verification, agent orchestration like the Ralph loop, and system evolution) to keep complex automations reliable and improvable. The speakers share practical insights on skills, hooks, sub-agents, and “second brain” setups, plus examples like Brain 2 in ClickUp and practical workflows across development, operations, and business processes.
Every Level of a Claude Second Brain Explained
The video outlines five levels of building an AI second brain, starting from simple file organization (level 1) and progressing to complex, autonomous systems (level 5). It covers how to structure data, routing, memory, semantic search, and knowledge graphs, with practical examples like claw.md, cloudmd, memory, and agent workflows, plus cautions on privacy and team adoption. The core message is to tailor your second brain to your actual data usage and pain points, rather than chasing a single “best” architecture.
We Might Actually Need to Stop AI
The video examines how Enthropic and OpenAI publicly advocate for slowing AI development through international coordination and verification, highlighting the tension between race-to-build incentives and safety measures. It contrasts expert-level calls for a treaty or global pause with broad public skepticism, explores how a future personal AGI could amplify power disparities, and强调 the difficulty of enforcing a slowdown given competitive incentives and the need for broad consensus.
Learn These 6 AI Skills Now (Before AI Replaces You)
AI is here to stay and will reshape many jobs, requiring you to future-proof by developing six core AI skills. The talk guides practical steps: become the AI person, hone taste and judgment, become a context engineer, maximize iteration speed, build autonomous AI systems (Jarvis), and create multiple income streams, all while keeping a clear North Star and personal motivation.
Claude Fable 5 Made This Entire Video By Itself.
The video demonstrates how Claude Fable 5 can autonomously generate a YouTube video, including script, voice, avatar, animation, and editing, and discusses the workflow, costs, and replicability. It showcases a fully AI produced production process and offers tips on prompting and verification.
From Zero to Head of AI in 1 Year (as a regular person)
In this interview, Eileen explains her new role as Head of AI across a group of 15 companies, outlining how she shapes AI strategy while staying hands-on in implementation. She links IBM’s survey findings about chief AI officers and employee readiness to her own career transition—from email development to AI leadership—emphasizing the importance of building in public, showing concrete proof of work, and fostering a culture shift to make AI adoption practical and non-threatening.
I Turned Claude Fable Into The Ultimate Second Brain
The video introduces Claude Fable as a powerful “second brain” and outlines how the creator uses it alongside Herk 2 to build an AI operating system built on four core elements. It explains the four Cs framework (context, connections, capabilities, cadence), emphasizes adoption, cost considerations, and best practices for building scalable, automated workflows with sub-agents while maintaining trust, ownership, and continuous improvement.
Claude Mythos is Finally Here.
The video analyzes the rollout of Claude Fable 5 and Claude Mythos 5, detailing pricing, how Fable 5 is included in subscriptions until June 22 and then requires credits, while Mythos 5 remains restricted through Glasswing. It compares capabilities to Opus, discusses safety and cyber security safeguards, and predicts how Mythos features may gradually be integrated into future Opus versions, with emphasis on practical usage and caveats for developers and enterprises.
How to Build Claude Subagents Better Than 99% of People
The video explains how Claude's Cloud Code sub-agents act as lightweight, parallel workers that can be delegated tasks without cluttering the main session. It covers how to create and configure sub-agents using YAML front matter (name, description, triggers), distinguishes sub-agents from skills, and shows how project/global scopes and dynamic workflows help manage research and tasks while preserving core context.
Get daily AI recaps from
Nate Herk | AI Automation in your inbox
Get AI-powered summaries delivered to your inbox. Save hours every week while staying fully informed.