Two types of AI companies: Why AI Process Companies will win
The most successful AI companies will be process-first
There will be two primary opportunities for startups in AI. The first is the skeuomorphic opportunity (mimicking what came before), while the second is truly AI-native, unlocking entirely new potential.
AI Product Companies versus AI Process Companies
The two types of AI companies are as follows:
AI Product Companies: Your company uses AI to build exceptional products, services, or experiences
AI Process Companies: Your company empowers users to use AI to produce exceptional products, services, or experiences, for themselves, with ease
For example:
An AI Product company will create awesome music using AI. The company will publish the best results and market them to consumers.
In contrast, an AI Process company will give users intuitive tools to generate their own music, customized to their taste and use case. Users will make a few simple inputs and music will be generated for them, on the spot, for their commute, run, deep work session, or dinner party.
While both companies will have their place in the future, AI Process companies will fundamentally disrupt how products, content, experiences, and services are created by abstracting the powerful AI models behind them, empowering users to generate incredible content at-demand.
AI Process companies will win long-term because:
Outcomes will be highly personalized for the users producing them, resulting in better content and higher enjoyment
AI Process companies will naturally evolve into social platforms, enabling users to share their generated outcomes which are never-before heard, seen, or experienced by other users of the same platform
Generated content brings the friction of content discovery to zero - you don’t need to scroll for hours looking for new music when it can be produced for you on the spot
When building an AI startup, consider whether it should be an AI Product Company or an AI Process Company - should you use the AI to produce outcomes for users, or should you allow users to use the AI to produce outcomes for themselves?
When you can, gives users tools, then empower them to share what they produce.
The easiest process with the best results, wins.