When Looking Professional Stops Proving Anything

A polished website tells you less about a business when almost anyone can produce one in an afternoon. The same might become true of a confident product description, a persuasive application or a detailed research report. Good presentation used to offer at least some evidence that someone had invested time, skill or money. As those appearances become easier to reproduce, we may have to work harder to find out what is behind them.

George Akerlof described a related problem in his work on the market for used cars. Sellers know more about a car’s quality than buyers do. If buyers cannot reliably distinguish a good car from a bad one, they become reluctant to pay a price that rewards quality. Owners of better cars may then decide selling is not worthwhile. The quality uncertainty can drive useful transactions out of the market. Everyone being free to buy and sell does not resolve the information problem.

That provides a way to think about an internet full of inexpensive, convincing material. Imagine an independent analyst who spends days checking the evidence for a report. Someone else generates a document that looks equally careful in a few minutes. If buyers cannot tell which one contains sound work, their willingness to pay for either may fall. The careful analyst can try to demonstrate the difference, accept a lower return or leave. The damage would come from the difficulty of identifying quality, even if excellent work remains available.

This is my application of Akerlof’s argument, rather than a claim that AI must produce poor work. It can help a careful analyst test an assumption or find a mistake. It can also give people with useful ideas the ability to present them clearly. The concern arises when the appearance of competence becomes much cheaper faster than the ability to verify competence. Better tools can improve the work itself while weakening the clues that customers previously used to judge it.

If that happens, a reputation built over time could become more valuable. So could a guarantee that carries a real financial consequence, or a result that another person can reproduce. A business might have to show how its product performs instead of relying on a convincing description. There is a less attractive possibility as well. Buyers may retreat to familiar names simply because checking an unfamiliar provider takes too much effort. Talented newcomers could face a higher barrier even while the cost of starting a business falls.

The cost of producing an impressive claim could therefore fall while the cost of establishing trust rises. I would expect that tension to shape how people hire, buy advice and choose information sources. A useful response would be to make evidence easier to inspect and responsibility easier to assign. Otherwise, we could end up with more professional looking material and less willingness to believe any of it. The businesses that help customers tell the difference would have something valuable to sell.