Nineteen AI predictions hit Colin Taylor's inbox. Not one mentioned the trap. Upstream, your expertise is locked in your head. Downstream, the platforms are consolidating under five companies rewriting terms quarterly. Both jaws are closing at the same time.
This one's landing on Sunday instead of Tuesday. Turns out "extract before you execute" applies to newsletters too. I wanted to give you this before the week starts, not after it's already running. New rhythm. Let me know if it lands differently.
Last week I said the standard doesn't change.
No matter how far the distance.
This week I almost broke my own rule.
Wednesday morning. Coffee's getting cold.
I'm four paragraphs deep into an industry newsletter.
19 AI predictions from founders, investors, operators...
I'm nodding along like everyone else.
Solo founders replacing 15-person teams.
Synthetic customers predicting purchase intent at 90%.
AI agents buying things on your behalf.
And I caught myself doing the exact thing I tell clients to stop doing.
Consuming someone else's vision of the future, instead of taking a closer look at what's actually breaking right now.
Nineteen powerful predictions.
But almost every single one pointed downstream.
Faster output. Better automation. More execution etc.
Not one really mentioned the trap.
Done well, AI solves the easy stuff.
First drafts, research, content at scale...that's pretty much a wrap.
But survey data from product leaders this month showed where the real gaps are widening.
It's not where you might think.
Not in writing. Not in coding.
It's in research. In strategy.
In knowing what to build and whether to build it at all.
The "whether or not" question.
Same one from last week.
While AI keeps getting better at the "doing".
The thinking/meaning layer keeps getting harder to reach.
And that thinking layer...
Your pattern recognition, your client instincts, your methodology that actually works.
It's still sitting in your head.
Homeless.
No structure. No documentation. No defensible form.
One of those 19 predictions broke this down into a moat framework I can't stop thinking about.
Tier 1 - Unique Data.
Think proprietary insights. Buyer psychology you've observed firsthand.
Decision frameworks proven over years.
AI can't (maybe shouldn't) touch this. Strongest moat there is.
Tier 2 - Process Knowledge.
How you do what you do.
Feels safe, until AI replicates process better every quarter.
Tier 3 - Execution.
Producing the work. Already a commodity. Zero moat.
Most of us think we're competing at Tier 1.
We believe our experience is the moat.
But experience that only lives in your head?
That's not Tier 1.
That's not even on the board.
And I promise you, this gap doesn't stay hidden.
That's the part nobody warns you about.
Every vague brief your team feeds ChatGPT that comes back generic?
That's your undocumented expertise making itself visible by its absence.
Every prospect who can't tell you apart from the AI-generated competition?
Same thing.
AI doesn't just ignore what you haven't extracted. It exposes it. Daily. To everyone.
Extract your expertise, your POV, or AI will prove you never had one.
If you're a consultant...
It's your methodology evaporating when you take a two-week vacation.
If you run a plumbing company with 15 trucks...
It's your best dispatcher's 22 years of routing instincts, gone the day he retires.
If you own a med spa...
It's the way your top aesthetician consults that closes 80% of her appointments.
Undocumented.
Unrepeatable.
Walking out the door the day she gives notice.
The moat only exists when it's extracted, documented, and structured.
Until then? It's memory.
And memory doesn't compound.
Doesn't transfer. Doesn't scale.
That's the first jaw of the trap.
Here's the second jaw.
And this one's moving fast.
While everyone argues about what AI can't do...
The platforms controlling what AI already does for your business are consolidating, repricing, and collapsing underneath you.
Last year, Builder.ai, $1.5 billion valuation, backed by Microsoft, went bankrupt.
Client applications?
Dark. Overnight. Business data? Inaccessible. Support? Gone.
Real businesses. Real owners.
Locked out of their own operations between Tuesday morning and Tuesday afternoon.
Industry estimates put the cost of recovering from a platform collapse like that at six to twelve months and hundreds of thousands of dollars.
And that's for a planned migration. Builder.ai wasn't planned.
Now look at what's happening this month.
OpenAI is projecting $14 billion in losses while testing ads inside ChatGPT.
They've acquired 10 companies since 2025.
Their API pricing can swing over 1,000x depending on which model and tier you use.
And those numbers can shift overnight.
And a survey released last week found 94% of IT leaders are concerned about vendor lock-in.
Ninety-four percent.
This isn't an enterprise problem.
It's the agency whose content workflow lives inside one tool that just got acquired.
The accountant whose client portal is on a platform that just jacked prices 40%.
The restaurant group whose reservations, reviews, and scheduling all sit on software that could be absorbed, repriced, or killed next quarter.
You're not renting tools. You're renting the right to operate.
That lease can be pulled the moment the landlord decides your rent isn't worth their margin.
Here's what I see and almost nobody is saying out loud.
Both jaws are closing at the same time.
Upstream, your most valuable asset, the thing that actually makes you different, is locked in your head.
Undocumented.
While AI commoditizes everything around it.
Downstream, the platforms running your day-to-day are consolidating under five companies who are burning cash, buying competitors, and rewriting terms of service quarterly.
Nineteen predictions hit my inbox this week. Not one mentioned this.
Your clients are drowning in those same predictions and they can't tell who's real anymore.
Every AI subscription you add sends money downstream to someone else's infrastructure while the one asset that costs nothing to extract sits untouched.
That's all four D.I.B.S. forces tightening the same trap.
Both jaws. Same root cause.
You haven't extracted what you own.
Your expertise.
Your methodology.
Your client intelligence.
Your proven systems.
The things no platform can revoke, no algorithm can replicate, and no competitor can copy, if they exist somewhere other than your head.
But here's what nobody on the other side of extraction tells you.
Once it's documented, AI stops being a threat and starts being a multiplier.
Every client interaction, every pattern you've proven, every decision framework feeds back into a foundation that gets stronger.
Your competitors are feeding AI nothing and wondering why they get nothing back.
Document your upstream, and AI amplifies what you've already proven instead of replacing what you never captured.
Build on your own foundation downstream, and platform shifts become annoying instead of fatal.
And what gets built in 90 days isn't a patch.
It's the infrastructure your business runs on from that point forward.
Tighter iterations.
Faster positioning.
Every tool you adopt, every person you hire, every client you serve, all of it works better because the foundation exists.
That's not a retreat from AI.
That's what makes AI actually work for you.
And understand, this is a compounding problem.
Every week your expertise stays undocumented, the competitors who ARE extracting theirs pull further ahead.
That gap doesn't hold still. It widens.
The advantage they're building isn't a lead.
It's an accumulating distance you can't close later by "catching up."
Two questions in the meantime. One for each jaw.
Upstream: If you handed your business to a stranger tomorrow, what percentage of what actually makes it work could they find written down?
Downstream: If your three most-used platforms vanished tomorrow, what percentage of your business still runs?
If either answer made your stomach drop, that's exactly what the March sprint is built to fix.
Both jaws. 90 days. And what comes out the other side isn't just protection.
It's a business that moves faster than it did going in.
If you're ready, book a strategy session.
Stay surgical,
Colin Taylor
Creator of The Asset Alchemy Method
Builder.ai bankruptcy: TechCrunch, May 2025
OpenAI projected $14B losses: The Information, Oct 2024
ChatGPT ads launch: OpenAI Blog, Feb 2026
OpenAI acquisitions: Tracxn, Jan 2026
94% vendor lock-in: Parallels 2026 State of Cloud Computing Survey, Feb 2026
Platform migration costs: Network World/Gartner, Feb 2025
What is the "two-jaw trap" closing on service businesses?
The two-jaw trap describes two forces closing simultaneously on business owners. Upstream, institutional knowledge and expertise remain undocumented while AI commoditizes everything around it. Downstream, the platforms running daily operations are consolidating under a handful of companies that are burning cash, acquiring competitors, and rewriting terms of service quarterly. Colin Taylor's D.I.B.S. framework identifies all four market forces accelerating this squeeze: Decision Fatigue, Inflationary Pressures, Buyer Bottlenecks, and Synthetic Content.
Why does undocumented expertise become a bigger liability as AI improves?
AI doesn't just ignore what you haven't extracted. It exposes it. Every vague brief your team feeds an AI tool that comes back generic reveals the gap between what you know and what's actually documented. Every prospect who can't tell you apart from AI-generated competition is seeing the same gap. The moat only exists when expertise is extracted, documented, and structured. Until then, it's memory, and memory doesn't compound, transfer, or scale.
What are the three tiers of competitive moats in the AI era?
Tier 1 is Unique Data: proprietary insights, buyer psychology observed firsthand, decision frameworks proven over years. AI can't replicate this. Strongest moat. Tier 2 is Process Knowledge: how you do what you do. Feels safe until AI replicates process better every quarter. Tier 3 is Execution: producing the work. Already a commodity with zero moat. Most service providers believe they're competing at Tier 1, but experience that only lives in your head isn't even on the board.
What happens when AI platforms collapse or reprice overnight?
Builder.ai, a $1.5 billion company backed by Microsoft, went bankrupt and left client applications dark overnight. Business data became inaccessible. Industry estimates put recovery from a platform collapse at six to twelve months and hundreds of thousands of dollars. Meanwhile, 94% of IT leaders report concern about vendor lock-in. Service businesses that build on their own documented foundation downstream experience platform shifts as annoying rather than fatal.
How do I start extracting institutional knowledge from my business?
Start with two questions. Upstream: if you handed your business to a stranger tomorrow, what percentage of what actually makes it work could they find written down? Downstream: if your three most-used platforms vanished tomorrow, what percentage of your business still runs? The Asset Alchemy Method addresses both jaws in 90-day sprints, extracting the expertise, methodology, client intelligence, and proven systems that no platform can revoke and no competitor can copy.
Ready to see what you're sitting on?
Book Your Diagnostic Call