I’ve spent fifteen years telling this audience the same thing: don’t listen to what an institution says, watch its incentives. That rule doesn’t stop applying just because the subject is AI instead of the Fed.
Most people still think about artificial intelligence in the most obvious ways — write an email, summarize a document, answer a question, write some code. Useful. But that’s the equivalent of buying a smartphone and only ever using it as a calculator.
The applications that actually matter for your financial survival aren’t the flashy ones. They’re showing up in places we don’t even think of as “computer problems” — friction, memory, coordination, leverage. And the honest question isn’t “what jobs will AI replace.” It’s “who captures the value when it does, and how do you make sure it’s you and not some institution with an incentive to make sure it isn’t.”
Here’s where I think the real leverage is.
AI as Your Argument Preparation Machine
This is the one I’d bet on first, because it’s the most directly useful to anyone reading this.
Before you negotiate almost anything — an insurance settlement, a contractor dispute, a medical bill, a parking citation, a property tax assessment — AI can prepare the battlefield. Upload the documents. It finds the inconsistencies, builds the timeline, calculates the numbers, and anticipates the other side’s arguments.
Not “write me an angry letter.” Something far more valuable: tell me where my leverage is.
I’ve built an entire legal strategy around exactly this kind of asymmetric research — a private citizen with a laptop doing work that used to require a firm. That’s not a hypothetical for me. That’s Tuesday. And it’s available to you now, for whatever you’re fighting.
AI as Corporate and Institutional Memory — Which Cuts Both Ways
Every institution has knowledge that lives in somebody’s head, or worse, in a filing cabinet nobody wants you to find. Ask why a city stopped enforcing an ordinance, why a bank restrained an account, why a claim got denied — and historically you got a shrug.
AI turns that opacity into searchable record. Emails, contracts, procedures, meeting minutes — an institution’s own paper trail becomes queryable.
Here’s the incentive-alignment part: the institutions with the most to fear from this are the ones whose current business model depends on you never finding the paper trail. Municipal parking systems, insurance claims departments, HOA boards — all of them rely, in part, on friction and confusion staying expensive to overcome. AI collapses that cost. That’s not a neutral technology story. That’s a power shift.
AI as a Bureaucracy Translator
Government forms, insurance policies, HOA covenants and corporate fine print are technically written in English the way a parking sign technically gives notice — which is to say, often not adequately at all.
AI can tell you plainly: here’s what they actually want, here’s the deadline, here’s where your filing is deficient, here’s the exact regulation they’re relying on. Millions of hours get wasted every year just figuring out how an institution wants something presented to it — and that confusion is not always an accident. Complexity is sometimes the product.
The Labor Market Question Nobody Wants to Answer Honestly
We keep asking what jobs AI will replace. Wrong question. The better one: what unnecessary human activity does it eliminate, and who was getting paid to perform that activity?
Scheduling. Searching. Reconciling. Following up. Explaining. Checking. None of these is anyone’s whole job — but they’re enormous chunks of millions of jobs, and entire layers of middle management and administrative overhead exist to perform them.
When that overhead compresses, the money doesn’t vanish. It moves. To whoever owns the tools, and to whoever was fastest to stop doing the eliminated task and start doing something AI can’t. The publicist stops scheduling and starts selling. The lawyer stops searching and starts arguing. The question for your portfolio and your career isn’t whether this compression happens — it’s already happening — it’s whether you’re positioned on the side that captures the gain or the side that gets compressed.
This is the frame I’d apply to every AI headline you read this year: not “is this impressive,” but “who gets paid less, who gets paid more, and which side of that am I on.”
