Should Startups Fine-tune Their Own AI Models?

Jon Noronha says early startups should not fine-tune or train their own models as the plan. Gamma has built on other companies' models from the start, moving among Anthropic, OpenAI, and Google. Training costs more than an early company can sensibly spend, and the models on the market keep getting better. He would rarely want to compete with them. The first job is real usage, and the models already out there are how a team gets it.

Fine-tuning comes after that usage exists. Gamma has used it to cut cost and, in some cases, to personalize. Judging whether a slide looks right has not gone well on the big models, so Gamma is doing its own research for that job. He still says to use the strong models that exist. Closed models, in his view, are underrated next to open source. The money is going to the closed ones because they are good, and Anthropic, OpenAI, and Google staying a few months apart keeps the pressure on an application company. Gamma's stack is broad but shallow: many models, used in simple ways. Prompts plus the right context, which he also calls context engineering, do most of the work, because an easy job can run on an off-the-shelf model. He puts 98% of the value on prompts. A prompt is easy, portable, and often written by Claude rather than by hand. Fine-tuning is the layer on top, for cost or the last 2% or 3% of quality.