Part 3: The Evolution
Chapter 8: What AI Is Eating
The Moment I Knew
December 2024. I was on a call with a founder who wanted help with his go-to-market strategy.
Halfway through our conversation, he said something that stopped me cold:
"I already ran this through Claude before calling you. It gave me a pretty good framework. I guess I'm just paying you to confirm I'm not missing anything."
He wasn't being rude. He was being honest.
And he was right.
The generic strategy advice I might have given him five years ago? AI does it better now. Faster. Free. Available at 3am when he can't sleep.
I sat with that for a week. Then I restructured how ANC works.
Because here's what I realized: the game changed. And most people are still playing by the old rules.
This Isn't Hype
Every few years, tech has a "this changes everything" moment. Most of them don't.
The metaverse didn't change everything. Web3 didn't change everything. Most "revolutions" are just hype cycles that enrich early speculators and leave everyone else confused.
AI is different.
I'm not saying this because I read it somewhere. I'm saying this because I'm living it.
At SimpleDirect, tasks that took us 6 hours now take 1. Code that required a senior developer now gets scaffolded in minutes. Content that needed a writer now needs an editor.
At ANC, the advice layer of our business — the part where someone pays us to explain frameworks and strategies — is getting compressed. Not eliminated. Compressed. The floor is rising. What used to be valuable is becoming table stakes.
This is happening right now. Not in five years. Now.
What's Getting Eaten
Let me be specific about what AI is commoditizing. Not theoretical. What I'm actually seeing.
Tools and software.
The SaaS playbook — build a tool, charge $29/month, get thousands of users — is dying. AI can build tools. AI can customize tools. The defensibility of "we built software that does X" approaches zero when anyone can prompt their way to a similar solution.
SimpleDirect started as a SaaS tool. We pivoted to services. Now we're pivoting again. The tool layer keeps getting thinner.
Information and expertise.
If your value is "I know things you don't know," you're in trouble. AI knows more. AI explains it better. AI is available 24/7 and doesn't charge $500/hour.
Courses, playbooks, frameworks, templates — all the information products that built the creator economy — are getting commoditized. Not worthless yet. But the floor is rising fast.
Advice and consulting.
Here's the uncomfortable one for me: generic consulting is dying.
If a founder can describe their problem to Claude and get a reasonable strategy back, why are they paying a consultant? The answer used to be "because the consultant knows things Claude doesn't." That's increasingly untrue.
The consultants who survive aren't selling knowledge. They're selling something else. I'll tell you what in the next chapter.
Commodity execution.
Anything that can be systematized, documented, and repeated — AI will do it cheaper and faster.
Writing first drafts. Generating code. Creating designs. Analyzing data. Summarizing documents. Scheduling. Research. Admin.
The "virtual assistant" job category is getting compressed. The "junior analyst" job category is getting compressed. Anything where the value was "I do the boring stuff so you don't have to" is under pressure.
The Pattern Recognition Trap
Here's what makes this so dangerous for people like us:
The skills that made you successful are now the skills that will destroy you.
Think about what got you here. For me, it was:
- Learning to code (so I could build products)
- Developing expertise in a niche (so I could charge for advice)
- Creating content (so I could build an audience)
- Building systems (so I could scale without hiring)
Every single one of those is getting commoditized.
Not worthless. But less defensible. The moat is filling in.
And here's the trap: your brain wants to keep doing what worked. Pattern recognition is how humans succeed. We find what works and we repeat it.
But when the game changes, pattern recognition becomes pattern imprisonment. You keep optimizing for a world that no longer exists.
I catch myself doing this constantly. "We should build a tool for that." "We should create a course on that." "We should systematize that process."
Then I stop and ask: Is this a layer AI can eat?
Usually, the answer is yes.
What I Got Wrong
I'll be honest: I was slow to see this.
In 2023, I was still thinking about SimpleDirect as a product company. Build the tool, get the users, charge the subscription.
In early 2024, I was still thinking about ANC as an expertise business. Know things others don't, package it into programs, charge for access.
Both of those models are under pressure now.
The pivot to direct mail services for SimpleDirect? That worked because direct mail has physical components AI can't replicate. You can't email a postcard.
But even there, the margins are compressing. The parts of the service that were "strategy" and "targeting" — AI does that now. What's left is logistics and relationships.
At ANC, we stopped selling information. We started selling something else: selection, judgment, accountability, stakes.
The founders who pay us now aren't paying for frameworks. They can get frameworks anywhere. They're paying because we chose them (selection), we make decisions with them about their specific situation (judgment), we're accountable for outcomes (accountability), and increasingly, we're taking equity instead of fees (stakes).
Those are the layers AI can't eat.
But it took me too long to see it clearly. I'm writing this chapter so you see it faster than I did.
The Timeline Is Shorter Than You Think
Here's what I believe about AI's trajectory:
2024-2025: AI matches or exceeds human performance on most knowledge tasks. We're here now.
2026-2027: AI agents coordinate between tools and complete multi-step workflows autonomously. A 5-person team does the work of 15-20.
2028-2030: AI handles most execution. The remaining human value is judgment, taste, relationships, and accountability.
I could be wrong on the timing. But the direction is clear.
If you're building something that depends on information asymmetry, expertise moats, or systematized execution — the clock is ticking.
What This Means For You
If you followed Part 2 of this book, you started from revenue. You found customers. You're solving problems for money.
Good. That's still the right foundation.
But the question now is: what layer are you in?
Are you selling information? (Increasingly commoditized)
Are you selling tools? (Increasingly commoditized)
Are you selling generic advice? (Increasingly commoditized)
Are you selling systematized execution? (Increasingly commoditized)
Or are you in a layer AI can't eat?
The next chapter tells you exactly what those layers are.
The Good News
I know this chapter sounds grim. It's supposed to. Because too many founders are sleepwalking into obsolescence, optimizing for metrics that won't matter in three years.
But here's the good news: the layers AI can't eat are more valuable than ever.
When information is free, curation becomes valuable.
When advice is commoditized, judgment becomes valuable.
When execution is automated, accountability becomes valuable.
When everything is available, selection becomes valuable.
When AI is everywhere, human presence becomes valuable.
The founders who understand this aren't scared of AI. They're using it to amplify the human layers that matter.
That's what Part 3 is about. Not how to hide from AI. How to position yourself where AI makes you more valuable, not less.
Let's talk about what survives.