Anthropic’s Mike Krieger advised startups to focus on domain expertise and customer relations.
Startups in law or biotechnology can offer value beyond large language models, he said.
Krieger suggested exploring new AI interfaces to avoid becoming simple AI model add-ons.
Trying to avoid becoming part of the “OpenAI killed my startup” meme? Anthropic’s chief product officer has some advice.
In an episode of “Lenny’s” podcast published on Thursday, Mike Krieger said three key qualities can help startups defend themselves from being taken over by AI giants and stay “defensible” for at least the next one to three years.
First, Krieger said startups with deep knowledge of areas like law or biotechnology can survive AI giants.
For example, he said that he doesn’t see Anthropic building a solution for scientific labs. “But I want that company to exist and I want to partner with it,” he said.
Second, Krieger said a good relationship with customers can help.
“Don’t just know the company you’re selling to, but know the person you’re selling to at the company,” he said, referring to advice from one of Anthropic’s product leads.
It’s also helpful to come from the domain yourself, or bring on a cofounder “who came from that world,” he said.
Krieger’s third suggestion was for startups to play with new AI interfaces.
“Do something that feels very advanced user, very power user, very weird and out there at the beginning, but could become huge,” he said.
Krieger cofounded Instagram in 2010 and was its chief technology officer until 2018. He later cofounded the news-aggregating app Artifact, which was shut down last year. He joined Anthropic last year.
“I don’t envy them,” he said about founders wanting to build in the AI space. “Maybe that’s part of the reason why I wanted to join a company rather than start one.”
Kriger did not immediately respond to a request for comment from Business Insider sent outside regular business hours.
Krieger’s comments are part of an ongoing debate in the startup community about how founders can avoid building AI “wrappers” — a dismissive term used to refer to simple applications that are built on top of existing AI models and can be easily offered by LLM companies themselves.
Last year, OpenAI CEO Sam Altman said his company would “steamroll” any startup building “little things” on top of its model. He said that companies who underestimate the speed of AI model growth risk becoming part of the “OpenAI killed my startup meme.”
Krieger has made several recent comments about the future of work and how people can AI-proof their careers.
In an interview earlier this year, Krieger said developers would be spending more time double-checking AI-generated code than writing it themselves.
Late last month, Krieger said that he still has “some hesitancy” about hiring entry-level workers and that Anthropic is focused on hiring experienced engineers instead.
Read the original article on Business Insider
