Will AI Kill Your Startup? The Question Investors Should Ask
Will AI Kill Your Startup? The Question Every Investor Should Ask
Investors, family office principals and founders who want a simple test for judging whether an AI company has a lasting business.
In their words
"The problem that we have to solve with each of these startups that we're looking at is as those premier models improve, does this company get better or worse?."
"If they are actually using the models and their product improves and the Anthropics of the world can't take over their business, that's the company to invest in."
"Everybody is AI. You don't see a pitch deck that doesn't say, "We are AI, we're going to change the world."."
What this talk covers
James puts one test to every AI startup: as the premier models improve, does the company get better or worse? He predicts that in 5 years, 80 to 90% of companies calling themselves AI will no longer exist. He adds that, in his estimate, probably 50% of SaaS companies that claim to be AI and already have revenue will be gone too.
He says a company that only stays the same as the models improve is not one he would write a check for. A company that uses the models, gets better, and cannot be absorbed by firms like Anthropic is the one to back. He applies the test to his own health tech company, which automates medical practice workflows. He asks the same question in his family offices, whether in Singapore, Africa or the United States.
From the source
- James predicts that in 5 years, 80 to 90% of companies that call themselves AI will not exist.
- He estimates probably 50% of SaaS companies claiming AI, with revenue today, will also be gone.
- His health tech company automates medical practice workflows that are customized down to the physician.
Sources
From Richard C. Wilson's Centimillionaire Strategies YouTube channel: Will AI Kill Your Startup? The Question Every Investor Should Ask. Quotes come from the video captions, lightly edited for readability. Figures are the speakers' own estimates and experience.
General education only. Not legal, tax or investment advice.
Five points worth keeping
- Every pitch deck says AI. James notes that almost no pitch deck lacks an AI claim. Saying "we are AI" no longer separates one startup from another.
- Ask if the company improves. The core question for each startup is whether it gets better or worse as the premier models improve. James calls it the number one question he asks.
- Staying the same means pass. If a company will only stay the same as the models advance, James says he would not write a check. Real gains from better models are what he looks for.
- Check if model makers can absorb it. The company to back uses the models and improves, and the large model builders cannot take over its business. OpenAI, Gemini and Anthropic are named as the model makers in question.
- Custom workflows resist takeover. In his health tech company, James says the whole medical practice workflow is customized down to the physician. A model may automate one step, but he does not expect it to take over the workflow.
Related questions
- Can AI score my pitch deck before I send it?Yes.
- What are the best AI-powered CRM tools for fundraising?There is no single best one.
- Can AI automate investor follow-up sequences?Yes, AI can automate much of the routine work in investor follow-up, like scheduled emails, reminders and tracking who replied.
Questions about this video
What question should investors ask about an AI startup?
James says to ask whether the company gets better or worse as the premier models improve. If it stays the same, he says he would not write a check.
How many AI companies does James expect to survive?
He predicts that in 5 years, 80 to 90% of companies calling themselves AI will not exist. He also expects about 50% of SaaS companies claiming AI, with revenue today, to be gone.
Which kind of AI company does James favor?
One that uses the models, sees its product improve, and cannot be taken over by the model makers. James gives his health tech workflow business as an example.
Put this into practice with the dilution calculator. See all ai for capital raising resources.
