Accept Change or Change Is Visited Upon You

The grammar is the argument

"You either accept change or change is visited upon you. And we're certainly in that moment." The line, attributed to financier Michael Milken, was carried forward by Apollo Global Management's CEO as the frame for the present AI moment.¹ "Visited upon" is passive and unavoidable: the maxim does not offer a choice between changing and not changing, since the change happens either way. The only variable is authorship. That is a narrower and more useful claim than the usual adapt-or-die exhortation, which implies survival is on offer to anyone who simply moves fast enough.

It binds hardest on winners

The application people tend to skip is the one that matters most: "we have to be established companies, including companies that are successful like ours, we have to be really paranoid about replacement risk." The mechanism is cost structure, not speed: "the cost of starting a new business, the velocity of starting a new business. We've seen more business startups than ever, because challengers now can start from a much different place."¹ An incumbent operates a business designed for yesterday's cost of software and data, embedded in headcount, process, and organizational structure, all of which are expensive to unwind, while a challenger designs for today's costs from a blank sheet.

A sharper internal diagnostic follows from the same logic: "everyone can envision how the job they do currently can change with the benefit of AI. A handful of people can actually envision, when data and software become free, how the business should exist versus how it does exist."¹ That gap, between improving the current job and re-deriving the business from scratch, is the difference between accepting change and merely absorbing it, and almost nobody, by this account, can do the second.

Tiredness as the correct state

The interview this line comes from returns repeatedly to exhaustion, framed not as a complaint but as evidence the posture is working: continuous paranoia about replacement is effortful in a way that never resolves, because there is no state of having fully adapted. The uncertainty about how the AI transition ends is itself the argument for staying paranoid rather than declaring victory.¹

What is not permanent

The claim is not that current products or structures are permanent, and the evidence offered is personal history: four instrument classes that now define fixed income (high-yield bonds, levered loans, ETFs, and securitization at scale) did not exist when Rowan's own career started in 1984.¹ "What got us over the past 10 years is not going to be what gets us where we need to go going forward. We are going to have to accept the markets have changed, that our role in these markets have changed, that the products that we currently think of as immutable are not immutable." The same person concluded he was "too late, all the money has been made" in 1984, again in 1990, and again in 2001, and was wrong each time because the business kept changing underneath the pessimism.¹