VCs love to talk about AI, but aren’t writing as many checks as you might think
techcrunch.comAnecdote from 1 VC is that they don't really know how AI companies can make money. It's like the leaked Google document that said that neither Google nor OpenAI have any moat in AI.
And this is Google. How is a basic startup supposed to make significant sums of money from AI when even Google doesn't believe it can create something that can't be replicated or leapfrogged fairly easily.
For all AI short of AGI, the value isn’t in the AI but rather in what you’re using it to accomplish…same as any other tool.
This is like asking “how do we make money from HTML”.
You don’t. You use it to build something else that people will pay you for.
The crucial part of building tools with AI though is how quickly it becomes obsolete by new models. A little different than other tools.
I don’t believe this to be true in the vast majority of tasks in which AI will be deployed. These tasks typically have boolean outcomes. A more sophisticated model isn’t better than a less sophisticated one that does the job.
And most AI is not and will not be the headline grabbing things we see today.
I mean, speaking as someone who's really worried about speed of AI research... that's good news to me.