But a business whose output is identical to everyone else's, because everyone is using the same models to solve the same problems, has no USP and no signal to customers to say why they're different.
The meme a while back about OpenAI having no moat? That's just as true for businesses depending on any public AI tool. If you can't find something that AI fails at, and also show this off to potential customers, then your business is just a lottery ticket with extra steps.
Most businesses don't compete on difference - most competitors are virtually indistinguishable from one another. Rather they tend to compete on brand identity and loyalty.
I think businesses assume the output of AI can be the same as with their current workflow, just with the benefit of cutting their workforce, so all upside and no downside.
I also suspect that a lot of businesses (at least the biggest ones) are looking into hosting their own LLM infrastructure rather than depending on third party services, but even if not there are plenty of "indispensible" services that businesses rely on already. Look at AWS.
Yes, GenAI content is cheap.
But a business whose output is identical to everyone else's, because everyone is using the same models to solve the same problems, has no USP and no signal to customers to say why they're different.
The meme a while back about OpenAI having no moat? That's just as true for businesses depending on any public AI tool. If you can't find something that AI fails at, and also show this off to potential customers, then your business is just a lottery ticket with extra steps.