The Deflationary Storm For Lawyers And Consultants

The Accountant’s Nightmare

There is a story, perhaps apocryphal, about a consultant who solved a client’s problem in fifteen minutes. He submitted an invoice for $10,000. The client, outraged, demanded an itemized bill. The consultant obliged:

Solving your problem: $500
Knowing how to solve it: $9,500

This joke has circulated through the professional classes for decades, a self-congratulatory parable about the value of expertise. What it actually reveals is something more fragile: the entire economic logic of professional services depends on the appearance of difficulty. The consultant cannot admit the work was easy. The lawyer cannot reveal that her junior associate spent forty hours researching a question she could have answered in four. The coder cannot confess that most of his “development time” was Googling Stack Overflow.

The game required a kind of strategic opacity. You were not selling hours—you were selling the mystique of competence, wrapped in the respectable fiction of time.

In 2026, we see that Artificial intelligence has no interest in playing this game.

The Collapse of the Knowledge Premium

The deflationary logic is merciless and mathematically simple.

For a century, knowledge work has operated on an implicit bargain: clients pay for access to trained minds that can navigate complexity they cannot. The scarcity of these minds—created through expensive education, professional licensing, and years of apprenticeship—justified the price.

But AI represents something unprecedented: a technology that replicates the performance of expertise without replicating its cost structure. The junior associate’s document review, the analyst’s research synthesis, the developer’s routine code—these were the hidden engines of professional leverage. Firms hired many juniors to do the grunt work, billed them out at multiples of their salary, and used the surplus to fund the partnership.

This pyramid is collapsing from the bottom up right now.

When AI can perform 80% of junior-level tasks at near-zero marginal cost, the entire economic architecture of “professional leverage” implodes. You cannot bill $400 an hour for work that a machine performs in seconds. Or rather: you can, but only until your competitor stops doing so. And someone always stops first.

The implications cascade upward:

For law firms: The billable hour becomes indefensible when “hours worked” is no longer a proxy for value created. How do you justify a time-based invoice when the AI drafted the contract in six seconds?

For consultancies: The research-and-PowerPoint machine that justified armies of analysts becomes obsolete. What remains is judgment—but judgment is difficult to meter and harder still to sell.

For software firms: The entire premise of “development time” dissolves when AI can generate, test, and refactor code in the space between your keystrokes. The client no longer needs your team; they need your architecture—and perhaps not even that.

But What if Speed Isn’t the Point?

The deflationary narrative assumes that the purpose of professional services is efficient production. Faster documents. Quicker analysis. More code per hour. The AI accelerates production, thereby reducing production costs.

But what if the most valuable work was never “production” at all?

How could the work of a lawyer really be defined? Her highest function is not document assembly; it is counsel—the slow, uncertain work of helping a client understand what they actually want, what risks they can tolerate, what their adversary is likely to do. This requires not just legal knowledge but human knowledge: reading the room, sensing hesitation, knowing when to push and when to wait.

Consider the consultant. His highest function is not research synthesis; it is translation—taking a CEO’s half-articulated anxiety and rendering it into a strategic question that can actually be answered. This requires spending time in ambiguity, tolerating confusion, and resisting the temptation to optimize prematurely.

And for the software developer, her highest function is not code generation; it is judgment—knowing which problems are worth solving, which technical debt to tolerate, which elegant solution will become an unmaintainable nightmare. This requires accumulated intuition that no language model, however capable, has lived.

The deflationary storm destroys the market for production. But it may, paradoxically, increase the market for wisdom.

The question is whether the professions can distinguish the two, or whether they are so habituated to selling production that they no longer remember what wisdom looks like.

The Future of Professional Services

Think about the historical moment for authors and the entire book industry after German goldsmith Johannes Gutenberg invented the printing press in Mainz, Germany, around 1440. Or the innovation process that began for clerks when Microsoft launched Excel, enabling spreadsheets to be created and formatted in minutes.

The firms that survive the deflation will be those that understand a simple truth: You can no longer sell hours because hours have become cheap. You can only sell outcomes, risk, and the irreducible judgment that machines cannot replicate.

Law firms, consulting firms, and software development companies will completely change their internal structures and business models from this year onward to survive. We see big names struggling at the moment, and an inevitable wave of layoffs in their administrative departments. A simple “change” workshop for the whole team does not fit here; this is a big system-wide change.

The pivotal strategy has to face these two main questions:

The old model asked: How much time did we spend?
The new model must ask: What change in the client’s condition did we produce?

Standardized offerings, AI-driven templating and review, and the emphasis on success fees are three basic components of the new work approach. A huge investment in platforms and data models, combined with a major organisational change to a new, structured workforce of senior AI experts, will be key factors for success from 2026 on. An AI-driven consulting platform offered under a subscription model, supported by human consulting, is one possible form of offer consulting firms will make in the future.

This shift has practical implications:

• Pricing moves from inputs to outcomes. Firms must learn to price risk transfer, success fees, and subscription access to expertise—not timesheets.
• Talent models invert. The pyramid gives way to the obelisk: fewer juniors, more experienced practitioners, and AI systems doing the first-pass work that once trained apprentices.
• Proprietary knowledge becomes the moat. If execution is commoditized, competitive advantage migrates to what the firm knows that others don’t—data, playbooks, accumulated case history, embedded client relationships.

But what happens to the people who trained under the old system?

The junior associate, the first-year analyst, the entry-level developer did the grunt work, but in doing so, they learned. They developed judgment by doing a thousand document reviews, running a hundred models, and writing ten thousand lines of code.

If AI replaces this basic work, where does wisdom come from?

The deflationary storm may solve the problem of productivity while creating a new problem of formation. We may end up with firms that are brilliantly efficient and entirely hollow. With machines augmented by a few aging partners who remember, dimly, what it meant to learn by doing.

We are building an economy optimized for speed, in which the one thing that cannot be accelerated—human maturation—becomes economically irrational. What happens when we have optimized away the process by which humans learn to be wise?