Chi‑Hua Chien, co‑founder of Goodwater Capital, says the AI model layer is already being commoditized and the firms that will profit most are those that embed AI into end‑user applications, not the ones that simply sell the models themselves.
He points out that the lag between cutting‑edge AI models and what can run locally on a phone has collapsed from roughly two years to six months, and he expects it to shrink to three months within the next year. This rapid convergence fuels hyper‑personalization, a trend he sees already in portfolio companies such as Midi Health, which uses AI to expand access to hormone‑replacement therapy for women, and entertainment startups that are scaling to hundreds of millions in ARR by tailoring experiences to individual users.
Chien predicts price competition will intensify—citing Google’s recent AI subscription price cut—as vertically integrated players bundle services for consumers. He expects investors to back businesses that turn AI into customized consumer experiences, while traditional infrastructure providers may see limited upside.
