I’m Building Something New With Michael Simmons
A quick heads up about the agentic AI cohort we’re launching on Monday.
Over the past year, AI Maker grew to more than 20,000 subscribers and became a Top 50 Substack in Technology, making it the fastest growing AI newsletter on Substack.
That did not happen because I found one magic prompt. In fact, there’s no magic prompt.
It happened because I started building a system around my work.
Behind my newsletter, I have source files, style guides, audience profiles, performance notes, draft archives, topic ideas, launch plans, and rules for how good work should look.
All of these are built to help me boost my productivity.
When I use an AI agent inside that system, it helps me do real pieces of the newsletter operation:
Read the right source material before giving feedback.
Compare a new draft against my actual writing style.
Turn messy ideas into clearer outlines.
Find gaps in a launch plan.
Repurpose one idea across newsletter, LinkedIn, Notes, and X.
Catch problems before something goes out.
I still make the judgment calls. I still choose the angle. I still do the final pass.
But the repeated setup work is no longer trapped in my head or scattered across my AI chats.
That’s the part I think more people need to understand.
The next shift in AI isn’t just better chat or better models.
The models are already good enough for a lot of serious knowledge work. The bottleneck is whether your work is set up in a way agents can actually use.
That is where the real productivity jump comes from.
Not from typing one clever prompt into a blank chat window.
From building an agentic workflow where AI can read your files, follow your standards, use the right tools, connect with the apps where your work already happens, and help execute bigger chunks of work.
Chat is still useful. I use it every day.
But if you only use AI as a chatbot, you are leaving most of the new capability on the table.
And this is why the shift is not easy.
Using chat AI is simple. You open a box, type a request, and wait for an answer.
Working with agents is different.
You need to understand:
Features: which agentic commands actually matter.
Planning: how to create a plan the agent can execute without getting lost.
Knowledge base: how to organize your files, chats, transcripts, and examples so the agent can use them.
Context files: how to teach the agent who you are, how you work, and what good output looks like.
Human in the loop: when to stay involved and when to let the agent run.
Connectors: how to connect the agent to the apps where your work already happens.
Error correction: how to find the root cause of AI mistakes so they do not keep repeating.
Evals: how to create checks so the agent can catch problems before they reach you.
Automation: how to turn repeated work into end-to-end skills and workflows.
Self-improvement: how to make the agent get better the more you use it.
Security: how to protect your files, data, permissions, and work.
Strategy: how to decide what to build when AI can build almost anything.
That is why learning agents is harder than learning chat.
Chat is a tool you use.
Agents are systems you design, guide, correct, and improve over time.
That is also why I do not think most people will make this shift just by watching a few tutorials.
They need structure. Examples. Support. And real work to build on.
That is why I’m building something new with Michael Simmons.
I first connected with Michael about a year ago, and we’ve been talking ever since about AI, writing, learning, and where this agentic shift is going.
Michael is one of the strongest long-form writers I know. He has spent years studying mental models, learning, and how to turn complex ideas into writing people actually understand. His work through Blockbuster Blueprint has reached 115,000+ subscribers, and he has been deep in AI-assisted thinking, Claude Code, and agentic workflows too.
Our strengths are complementary.
So we decided to build something together.
It is called Agentic Academy for Knowledge Work.
It is a 10-week cohort for people who want to move from AI chat into agentic AI work.
The cohort starts Monday, June 15 at 11:00am EST.
I’ll share more on next Monday, but I wanted to give you the heads up first.
If you’ve been feeling like AI is useful but still too manual, this is the shift we’re going to help you make.
If you want to keep up with the launch notes and the full announcement next week, subscribe to Agentic Academy for Knowledge Work here:
More Monday,
Wyndo
 | A guest post by
|
Massive congratulations, that kind of growth in a single year is no small thing. The systems point is what most people miss, they go chasing one clever prompt when the real work sits in the unglamorous scaffolding that tells the tool what to aim for.
Most people are one good system away from a real productivity jump.