How do we price/value design work when the artifact is nearly free but the judgment behind it isn't?

The shift

Generating a polished, plausible design artifact — a mockup, a full comp, a variant set, a rebrand exploration — goes from scarce (skilled hours) to abundant (seconds, near-zero marginal cost). The correlation clients used to rely on — "a polished artifact implies good judgment produced it" — breaks, because the same tools now produce polished artifacts regardless of the judgment behind them.

The axioms

• Clients pay by hour or day-rate because producing a polished artifact took skilled labor-hours that were genuinely scarce.
• Iteration/revision rounds are priced as extra work because each additional version cost real designer-hours to produce.
• A designer's portfolio of past artifacts is the primary evidence of their judgment, because artifacts were the only externally visible trace of an otherwise invisible thinking process.
• Clients evaluate design work by the polish of the deliverable, because polish was expensive enough that only real skill reliably produced it.
• Agencies scale revenue by adding billable designer-hours — headcount is capacity.
• Someone must decide which of many plausible directions is right for this business, this brand, this moment — a judgment call, not a production task.
• Someone is accountable when a design decision fails in market — conversion drop, brand damage, an accessibility or legal miss.

Invalid axioms

1. Pay by the hour for artifact production. Time-to-produce-a-comp was the scarce input the fee was really metering. AI collapses that time toward zero, so hourly billing on production work prices a cost that no longer exists. Habit-trap: agencies and freelancers still quote based on expected hours-in-Figma, then pad hours to preserve margin as the real work shrinks — clients will notice the mismatch before the field admits it.
2. Revisions are billable extra work. A tenth variant used to cost as much designer-time as the first. Now variants are nearly free to generate, so metering revisions as scarce labor prices something abundant. Habit-trap: change-order fees and "revision caps" in contracts, written for a world where iteration was expensive.
3. Portfolio-of-artifacts as proof of judgment. A slick portfolio used to be reliable evidence someone could produce slick work, because slick work was hard to fake cheaply. Now anyone can generate a polished portfolio piece in minutes, so artifact quality stops being a good signal of the person's judgment. Habit-trap: hiring and vendor selection that still screens primarily on deliverable polish.
4. Headcount as capacity, capacity as revenue. Agency economics assumed more billable hands meant more billable output. When output generation is cheap per person, revenue scaling by headcount growth breaks down — the constraint moves elsewhere. Habit-trap: agencies still pricing retainers and pitching growth in terms of team size rather than judgment throughput.

Unchanged axioms

1. Someone must decide which direction is right for this business, this audience, this moment. AI can generate a hundred plausible directions; it has no stake in the brand, no read on the CEO's risk tolerance, no sense of what this specific market will forgive or reward right now. That's judgment on novel, high-stakes ambiguity — the thing the lens says stays scarce.
2. Someone is accountable when the design fails in market. If a rebrand tanks trust, or a checkout flow buries a legal disclosure, a human with a name and a contract is on the hook — a model isn't liable and can't be. Accountability doesn't get cheaper just because drafts do.
3. Trust built across repeated engagements. A client keeps hiring the same designer or agency because judgment proved reliable under real stakes over time, not because any single deliverable looked good. That track record is relational and slow to build precisely because it can't be generated — it has to be earned.
4. Taste — knowing what's worth making at all. Generating variants is abundant; deciding which problem is worth solving, which of the hundred directions deserves the client's money and reputation, is a goal-setting act AI doesn't do on its own initiative.

New axioms

1. If artifacts no longer signal judgment, what does? Clients lose their cheap proxy (polish) for evaluating a designer before hiring them, and have no cheap replacement yet — the field hasn't built a way to price or vet judgment directly, ahead of the outcome.
2. How do you price something invisible until it's tested in market? Judgment's value shows up in outcomes (conversion, retention, brand equity) that lag the design decision by weeks or months. Pricing on time-spent used to approximate this; once time-spent stops correlating with quality, there's no fast substitute metric.
3. Who absorbs the cost of AI-generated volume that still needs judgment applied to it? If a junior or a client can generate fifty options in an afternoon, someone still has to evaluate, cull, and defend a choice among them — that evaluation labor was previously bounded by how few options existed to begin with. Abundance of drafts creates a new, unbudgeted evaluation bottleneck.
4. How do junior designers build judgment when the artifact-production reps that used to train it are gone? Judgment historically got built through thousands of hours producing artifacts and absorbing feedback. If that production work is now automated away, the apprenticeship pipeline that grows senior judgment has no clear replacement.

Where it breaks

Billing by the hour for production (invalid) collides directly with judgment's value only showing up after the fact in market outcomes (new): clients will refuse to pay hourly rates once they can see the artifact took minutes to generate, but nobody has a working alternative — outcome-based fees, retainers for judgment, or value-based pricing — that can be verified fast enough to invoice against. The field is stuck between a pricing model that's visibly obsolete and a replacement that isn't built yet.

A second collision: portfolios stop proving judgment (invalid) right as the market most needs a fast way to vet judgment before hiring (new) — the very moment clients need a better signal, the old signal breaks and no cheap substitute exists, so hiring and vendor selection get noisier, not clearer.

ai-axioms.vercel.app/axioms/product-design/design-pricing-judgment-ai-axioms