The World That Comes After

I feel like I can see the future.

Not perfectly. Not the companies that win, the models we use, or what we’ll call any of this. But I think I can see the direction.

For most of the digital era, we have adapted ourselves to technology. We learned the interfaces. Opened the apps. Navigated the menus. Searched for the information. Clicked the buttons.

I remember hearing the idea of ambient technology almost a decade ago — technology that gradually disappears into the environment around us. I think AI may finally make that idea real.

McKinsey recently discussed part of this transition as moving from “clicks to conversations.” I think it goes further than that.

Clicks → conversations → ambient intelligence.

Instead of learning how to operate technology, we increasingly just express intent. We speak. We ask. We delegate. Eventually, perhaps, we barely notice the technology at all.

It understands the context, coordinates the systems underneath it, and acts.

That is the puck I’m trying to skate toward.

Because eventually, “AI” itself may become ordinary. Invisible. Just part of how work gets done. Just… tech.

And that is exactly why now matters.

Right now, I happen to be watching that technological transition collide with three enormous forces in financial services:

• $124T of wealth is expected to change hands through 2048. (Cerulli Associates)

• 105,887 advisors — 37.4% of the industry — are expected to retire over the next decade, representing 41.4% of industry assets. (Cerulli Associates)

• AI is entering a historic capital cycle: Goldman Sachs estimates $765B of AI CapEx in 2026, rising to $1.6T annually by 2031. (Goldman Sachs)

Think about what happens when those curves intersect.

Capital is changing hands. The people managing it are changing. And the interface between humans and institutions is changing with them.

Voice AI. Agents. Ambient computing. Autonomous workflows.

These may look like separate technology trends today. I increasingly think they are pieces of the same transition.

Technology is moving from something we operate to something that operates around us.

Imagine something much more ordinary.

A client gets promoted, their compensation increases, they are expecting a child, and they begin thinking about buying a home.

Today, each of those events might live in a different place. Payroll knows about the raise. A benefits portal knows about the new compensation package. A calendar knows about the parental leave dates. A bank sees the growing cash balance. A brokerage account sees the portfolio. A mortgage lender eventually sees the home search. The financial advisor may only learn the full picture months later, during a scheduled review.

In the world I see, the system already understands that these events belong together.

With permission, the client’s financial environment could recognize the promotion, update expected cash flows, adjust savings targets, optimize retirement and tax elections, model the cost of a larger home, increase emergency reserves, update insurance needs, revisit the estate plan, recommend how much equity compensation to hold or diversify, and automatically redirect excess cash toward the client’s highest-priority goals.

If the client begins looking at homes, the system could understand what price range is actually sustainable before they ever speak to a lender.

If markets fall, it could recognize that the down-payment horizon has shortened and reduce risk without waiting for the next quarterly review.

If the baby arrives early, spending changes, or the promotion is delayed, the plan changes with it.

The client does not have to open six applications, reconcile six accounts, or remember to tell every institution that their life has changed.

Their financial life simply adapts around them.

And the advisor is still there.

Not to manually move money from one bucket to another.

But to ask whether buying the larger home is actually the life they want.

To challenge the system when the financially optimal decision conflicts with the family’s priorities.

To help them think through tradeoffs that cannot be reduced to an optimization problem.

To make sure the technology is serving the life, rather than quietly designing it.

That is what I mean by ambient intelligence.

The client should not have to constantly operate their financial system.

The financial system should increasingly understand the life being lived around it.

Now imagine something far more difficult.

A client receives devastating news from a doctor: they have a limited amount of time to live.

Today, that might trigger a series of separate phone calls, appointments, forms, portfolio reviews, estate-planning conversations, tax consultations, and decisions made across disconnected systems over weeks or months.

In the world I see, the event itself could become the trigger.

With the client’s permission, their health data could communicate a material change in circumstances to their financial system. An agent could immediately understand the implications across their entire financial life: portfolio risk, liquidity needs, tax exposure, estate documents, insurance, charitable intentions, beneficiaries, required distributions, debt, and the timing of every future cash need.

It could model the remaining years of their life, continuously optimize the portfolio around those needs, identify the most tax-efficient sequence of distributions, surface estate-planning actions, prepare the documents that require signatures, coordinate with attorneys and accountants, and adapt as markets, health, spending, or family circumstances change.

And yet the financial advisor may become more important, not less.

The advisor would not spend the meeting manually calculating whether the portfolio should be 60/40 or 55/45. The machine could do that continuously.

The advisor would sit across from the client and their family.

Explain what is happening.

Challenge an assumption the system cannot understand.

Recognize that the mathematically optimal decision is not necessarily the humanly correct one.

Help a spouse prepare.

Help children understand.

Ask what the client actually wants the remaining years of their life to mean.

And ultimately verify, approve, and take responsibility for the actions the technology recommends.

That, to me, is the financial advisor–AI intersection.

Not a chatbot bolted onto today’s wealth-management model.

A continuously intelligent financial system operating around the client, with a human being still sitting at the point where judgment, trust, empathy, and accountability matter most.

The machine manages the continuous complexity.

The advisor manages the human meaning.

And financial institutions are an incredible place to watch that transition happen because the constraints are so unforgiving.

Money has to be protected. Customers have to trust. Regulators have to understand. Systems have to work. And someone ultimately has to be accountable.

So being inside this transition now means getting to watch institutions decide:

What can a machine know?

What can it say?

What can it do?

What should remain human?

What happens when the interface disappears but responsibility doesn’t?

Those questions feel much bigger to me than AI.

They are questions about how institutions — and eventually humans — operate when intelligence becomes embedded in the environment around us.

And that is why I don’t think the goal is to become an “AI person.”

Models will change. Interfaces will disappear. Today’s breakthrough will become tomorrow’s infrastructure.

But the more capable the technology becomes, the more convinced I am that the scarce things become deeply human:

Judgment. Taste. Trust. Courage. Clarity.

If intelligence becomes abundant, discernment becomes more valuable.

So I’m trying to skate past the technology itself. Toward the world the technology creates.

And professionally, that is the person I am trying to become.

Not merely the operator today’s institutions require, but the operator-executive tomorrow’s institutions will seek.

I don’t think we fully know what that person looks like yet. Their skills, responsibilities, authority, and judgment will be revealed as the technology matures and institutions reorganize around it.

But I suspect the people who eventually lead that world will not be the ones who waited for the role to become obvious.

They will be the ones who were already moving toward it.

Learning how intelligent systems actually get adopted. How operating models change. How governance evolves. How human judgment fits around automation. How trust survives when machines begin acting on behalf of institutions.

I’m trying to skate toward the operator-executive role that does not fully exist yet. The Master Operator-Executive.

By the time everyone can clearly see where the puck is going, the advantage may already belong to the people who have been skating there for years.

To be close enough to this transition that, when AI becomes ordinary, I have already learned how institutions change, how capital moves, how humans and machines divide responsibility, how decisions get made, how trust gets built, and how people behave when the ground beneath them starts moving.

Maybe that is the opportunity of this moment.

Not merely learning the technology.

Learning the world that comes after.