> Seseri made it clear that just because you connect to some AI APIs, it doesn’t make you an AI company. “And by AI-native I don’t mean you’re slapping a shiny wrapper with some call to OpenAI or Anthropic with a user interface that’s human-like and you’re an AI company,”
I personally agree with this perspective, but I think it runs contrary to how many companies new LLM-API-powered companies are choosing to present themselves. More than once I've seen marketing talking about "advanced machine-learning algorithms", when it's very clear by the single person engineering team that they're really just hitting an endpoint.
Not to say these aren't real products or real companies! But there's certainly a distinction between a startup powered by OpenAI and one powered by its own novel model.
To provide a counterpoint: were companies in the early days of the word wide web that started using the web to enable people do do things online that they couldn’t before not web startups?
All they were doing was using the technology innovations of other companies (browsers, ISPs, etc) applying a thin slice on top that targeted them at a particular problem.
I see the AI/LLM world as much the same. You’ve got some companies building the infrastructure (OpenAI etc) that enables others to layer bits on top to solve human/business problems.
I guess the argument is that “couldn’t the LLM just do the whole thing on its own”, but the reality of that is no.
I personally agree with this perspective, but I think it runs contrary to how many companies new LLM-API-powered companies are choosing to present themselves. More than once I've seen marketing talking about "advanced machine-learning algorithms", when it's very clear by the single person engineering team that they're really just hitting an endpoint.
Not to say these aren't real products or real companies! But there's certainly a distinction between a startup powered by OpenAI and one powered by its own novel model.