AI Is Repricing Financial Advice. What Will Clients Still Pay Advisors For?

Written by: Nyle Bayer

Financial advice is being repriced right now, and most of the industry is charging for the part that's headed to zero.

Somewhere in the depths of your CRM right now is your best future client.

She's 40 years old. She has a good income, a mortgage, kids, equity comp she doesn't fully understand, and about ten more years of compounding before her balance sheet makes her a qualified prospect for every firm in this industry.

And this morning, before she had coffee, she used AI six times. It summarized her inbox. It rewrote a memo. It answered a question about her health plan. It planned her kid's birthday party. She doesn't call it artificial intelligence anymore. She calls it "hang on, let me ask."

In ten years, that woman walks into a wealth management firm. Maybe yours, hopefully mine, and she is going to ask herself the same question every client has always asked, just with a new comparison point.

What am I paying you for that I can't get from the thing I already talk to every day?

I gave a version of this piece as a talk on the main stage at Future Proof last week. I spent four years helping build that event as its chief marketing officer, and this May I went back to running a registered investment advisor, Up Capital Management in Northern California. What I spoke about from that stage is a huge portion of what compelled me to go back to RIA-land. It's my attempt to give an honest, evidence-based answer to her question above. Not a prediction about where AI is going, a description of what is being repriced right now, this year, in the data, and what it means for how we build our firms.

This is the fourth repricing, not the first

Every one of the previous three followed the same script, so it's worth learning the script.

In 1800, three out of four American workers farmed. The valuable knowledge of that era lived in your hands and your back. The scarce resource was labor, muscle and hours, and at harvest the hands were the bottleneck on everything.

Then the disruptors arrived, and they have names and dates. Cyrus McCormick patented the mechanical reaper in 1834. Each machine that followed learned the codified part of farming, the repeatable part, and one farmer with a reaper did the work of five with scythes. Muscle became abundant. By 1900 farming was down to 41 percent of the workforce, by 2000 under 2 percent. We didn't stop eating. We produce more food than ever. The work didn't disappear. The premium moved.

And where did it move? To the factory, where the new scarce resource was capital. Then in 1952, engineers at MIT demonstrated the first numerically controlled machine tool, punch tape that could replicate a master machinist's precise movements without his years of accumulated feel for the work. One year later, in May of 1953, manufacturing's share of American employment hit its postwar peak. The disruptor arrived one year before the top.

So the premium moved again, to the desk, and the scarce resource became knowledge itself. Peter Drucker named us in 1959, knowledge workers. In 1940 fewer than 5 percent of American adults held a college degree. Today it's pushing 38 percent, and the price of acquiring one grew faster than nearly anything else in the economy. That is what a scarce resource looks like while it's scarce, everyone pays anything to get it. For seventy years the deal held. Learn scarce information, get certified in it, charge a premium for access to it. Some of us added a fourth step and put the letters after our name on LinkedIn. I have two sets myself, so I say this with love.

The pattern is ruthlessly consistent. When a new force makes the previous era's scarce resource abundant, the premium that scarcity commanded collapses, and the people who figure out what the next scarce resource is, and move toward it before the majority does, capture an asymmetric advantage. We should not take this as a threat by the way, every one of these transitions produced a world where a wider range of human capacity got recognized, developed, and paid.

Here's why this one is urgent rather than interesting. McCormick's reaper had to be manufactured in Chicago, shipped by rail, and bought one farm at a time. It took a century to remake the workforce. AI improves through a software update that reaches every system running it overnight. The farm shift took 150 years. The factory shift took 50. This one is showing up in national labor data within three.

The distinction that explains everything

The Federal Reserve Bank of Dallas published a framing this year that I think is the single most useful lens for our industry. There are two kinds of knowledge in every professional job.

Remember the machinist? In 1952 his accumulated feel was tacit knowledge, right up until engineers codified it onto punch tape. Every era's automation works the same way, it finds the part of the craft that can be written down and writes it down. The question for us is which parts of our craft are punch-tape-able.

Here's what the Dallas Fed found, and it's the sentence I'd tattoo on the industry if I could. AI replicates codified knowledge but not tacit knowledge. It substitutes for the textbook layer of a profession while it complements the experience layer. Same job, same year, one layer repriced down, the other repriced up.

This isn't a forecast. It's in the employment data already.

Same occupations. Down 6 for the codified layer. Up 6 to 9 for the tacit layer. That isn't job destruction. That's a repricing, happening inside job titles, right now. The Stanford authors call these early indicators, not causal estimates, and I'd say the same. But the wage data points the same way. PwC analyzed over a billion job postings this year across 27 countries and found that roles where AI amplifies the experienced practitioner are seeing twice the job growth and 42 percent faster salary growth than roles where AI made the core skill accessible to everyone.

The market has already voted. It is discounting textbook knowledge and paying a growing premium for judgment. The only question left is what that means for an industry that spent seventy years selling, certifying, and billing for textbook knowledge.

Run our own profession through it

Layer one, portfolio construction and rebalancing. Already repriced. That happened a decade ago. We all spent 2015 panicking about robo advisors, then we hired them. They live in our tech stacks now, we call them rebalancing software. Nobody wins a client because their pie chart is better (if that was the case I would win all the clients).

Layer two, analysis and plan production. The Monte Carlo, the retirement projection, the Roth conversion math. We used to hand clients the plan in a leather binder heavy enough to double as home security. I know firms that have phased out paraplanning roles entirely, replaced by AI workflows, meeting prep that took four to six hours now takes under one. The plan didn't get worse, it got abundant, and nothing abundant commands a premium.

Layer three, process and service. Notes, follow-ups, onboarding, the operational connective tissue. A majority of registered investment advisors already use AI tools here. Within a few short years this layer is invisible plumbing, like email.

Layer four, judgment, trust, and accountability. Knowing the client. Sitting with the widow. Talking someone out of selling everything in March. Being the human whose name is on the outcome.

That layer is appreciating. It's the only layer appreciating. And here's the uncomfortable audit every one of us should run this quarter. Look at your own fee, your own marketing, your own website, and ask what percentage of your stated value proposition lives in layers one through three.

For most of this industry, the honest answer is most of it. We describe ourselves by our artifacts. Go read your own website tonight. If it says "comprehensive, holistic, personalized planning," so does every other website in the industry, and more importantly, so does the chatbot. We all copied the same homework. Every one of those words is codified. Every one of them is on the wrong side of the repricing.

The paradox nobody says out loud

Look at how we charge. Basis points on assets. Our fee is literally attached to layer one, the portfolio, the first layer that got repriced. We price ourselves on the thing the machines took first, and we hand out the judgment, the only appreciating layer, for free at every review meeting.

I want to be careful here, because this is where most versions of this argument overreach. The easy conclusion is "so move your fee to judgment." I don't think anyone has earned the right to say that yet. Clients have never written a check for judgment at scale. They've written checks for portfolios, and judgment came bundled. Whether a buyer exists for the unbundled version, at what price, in what form, is an open problem, and I don't think anyone has solved it.

But the paradox is real regardless of how it resolves. An entire industry's revenue is indexed to its most commoditized layer while its most valuable layer has no price at all. That's not a stable arrangement. It's the kind of gap that either gets closed by the incumbents or filled by someone else.

The good news is the part that's left is the part that mattered all along. Clients never actually paid us for documents. They paid us for the feeling of "someone I trust is accountable for this." AI didn't invent that truth. It just stripped away everything we were hiding it behind.

The front door problem

I know what some of you are thinking, because I'd think it too. Everyone has been saying the human connection is the answer for two years. This isn't new.

You're right. "Be more human" is the consensus at every AI panel in this industry, and consensus advice, by definition, is not an edge. But here's the question nobody on those panels answers. If human connection is the product, where does the connection start? Where does the client come from?

Every lead magnet in wealth management is the same move, give away codified knowledge as bait to prove expertise. Our entire prospecting playbook is built out of the exact layers that just went to zero. "Be more human" is advice about the middle of the relationship. Nobody is talking about the beginning of it.

So that's what we rebuilt at Up. Not the middle. The front door. One firm's honest experiment, not a proven playbook. And rebuilding it forced us into a question much older than AI.

What is wealth?

Our industry has used that word as a synonym for money for a hundred years. Wealth management means money management, everyone knows that's the real translation. And that was never a philosophical choice, it was an economic one. Money is the measurable part of a life, the codifiable part, the part you can standardize, put on an exam, and bill basis points against. We built an entire industry on the codified fraction of our own word. And that fraction is the one being repriced.

Wealth never meant money. The word comes from the Old English weal, wellbeing, the same construction as health from heal. Wealth is the condition of a whole life going well. Money is a tool inside it, the way sleep is a tool inside health, essential, and not the same thing as the thing. We didn't redefine wealth at our firm. We gave the word its definition back.

And it's the same split we've been tracing the whole way though this article. Money advice is codifiable, which is exactly why machines can do it now. Advice about a whole life, whether it's well designed and whether the money is serving it, is tacit. It requires a human who can see another human. The true definition of wealth and the location of the surviving premium turn out to be the same place. That isn't a coincidence. That's the repricing telling us what our job was all along.

So if analysis is abundant, what's still scarce at the top of a funnel? Being seen. Not being analyzed, being seen. So, instead of building a lead magnet we built an instrument configured around a question no calculator has ever asked a prospect. Not "how much do you have," but "is your one life actually well designed, and is your money serving it." It takes the recovered definition and makes it measurable, four dimensions of a financial life, how it's structured, how it's experienced, what it's anchored to, and what it's for. It produces a reading of design, deliberately not a portfolio grade. The first thing a prospect experiences from our firm is not our expertise. It's a moment of recognition about themselves (more on this next week).

AI didn't replace the advisor at our firm. It relocated the advisor. The machine does the first pass of seeing. The human does everything after.

What kind of firm this requires

A firm built for the appreciating layer needs a different mix of people than a firm built for artifacts. I've spent the last year figuring out what that mix is, and I'm not done, but four roles keep showing up.

The operator. Someone who has run the inside of a wealth management firm and is tired of running it around deliverables. Who can take a project from idea to finished without a committee. Who has seen how the industry actually works and wants to fix it from inside a firm rather than complain about it from a panel.

Advisors who want to be paid for judgment. Experienced practitioners who already know their value is layer four and are done hiding it behind plan production. The people who have sat with two hundred families and know which technically right answer is wrong for this one.

Apprentice advisors. The entry-level rung is the one being repriced hardest, and this industry already faces a massive succession problem with a large share of advisors retiring in the next decade. If AI does all the grunt work juniors used to learn on, where does the next generation's tacit knowledge come from? We want early-career people who'd rather get real reps in real client conversations in year one than spend four years producing plans a machine now produces.

A builder. Someone who can turn a framework for seeing a client into a delivered experience. Part product, part operations, fluent in AI tooling and comfortable inside a fiduciary shop where the compliance answer matters as much as the design answer.

Every one of these is someone from inside wealth management who has real operating experience and has stopped believing the industry as it stands is the best version of itself.

The ask

Let's go back to her. The 40-year-old, six AI conversations before coffee.

In ten years she's your qualified prospect. And she will not pay you for a financial plan. She can generate one on her phone, in the parking lot, before the meeting, and it will be good. She'll pay for the thing she cannot generate. A person who has sat with two hundred families like hers. Who will pick up the phone at the worst moment of the next bear market and say "I know, and here's what we're going to do." Who is accountable, by name, for the outcome. Who knows what her wealthy life looks like and how money can best be put to work to serve it.

Muscle got repriced, and we kept eating. Hands got repriced, and we kept building. Now textbook knowledge is getting repriced, and people will still need advice, more than ever, because abundance of answers is not the same thing as clarity.

The premium always survives. It just moves to whatever stays scarce.

True to everything above, there's no lead magnet here (if you know me, you know I've thought lead gen died a long long time ago). Nothing to download, no funnel, no free guide. If you're already headed in this direction, connect with me and let's compare notes. And if you read the four roles above and recognized yourself, I'm not hiring in the usual sense, I'm looking for the right people to build this with. Reach out. Either way, good luck out there.

Sources

• Historical agricultural labor share, EH.net and Carter et al., Historical Statistics of the United States
• Manufacturing employment share, Bureau of Labor Statistics, Current Employment Statistics
• Knowledge worker share, Wolff, decennial census analysis, Economic Systems Research
• College attainment, US Census Bureau, Educational Attainment historical tables
• Federal Reserve Bank of Dallas, codified vs tacit knowledge framing, February 2026
• Brynjolfsson, Chandar, and Chen, "Canaries in the Coal Mine," Stanford Digital Economy Lab, August 2026 revision
• PwC, 2026 Global AI Jobs Barometer
• Etymology of wealth, Oxford English Dictionary and Etymonline

Related: The Exit You Want Isn’t Necessarily the Exit Your Business Is Built For