Goodwater’s AI thesis says the best AI companies may stop selling AI
Goodwater’s Chi-Hua Chien argues that the biggest AI winners may be consumer and prosumer companies where AI becomes invisible infrastructure rather than the product being sold.
The most successful AI companies may not scream “AI” in their branding. They may simply use AI to make the product feel faster, smarter and easier.
What happened
Goodwater’s Chi-Hua Chien argued that some of the biggest AI winners may not sell AI as the product.
The thesis is that as the model layer becomes more commoditised, value may shift toward consumer and prosumer applications where AI works quietly in the background. In that version of the market, the user buys the outcome, not the technology label.
Why it matters
This is a useful counterpoint to the current AI startup wave, where many companies still lead with “AI-powered” as the main pitch.
But consumers usually do not care whether a product uses a foundation model, a workflow engine or a clever interface. They care whether it saves time, feels magical, solves a real problem or becomes part of their daily routine.
The bigger picture
The next AI winners may look less like model companies and more like excellent product companies with AI inside.
That matters for founders because the moat may not be the model itself. It may be distribution, workflow ownership, user trust, brand, data loops and product taste. In other words: AI becomes the engine, but the company still needs a real product.