That’s my read too, if I’m being generous.

I also think OP is missing the point saying the target applications are too small of a market to be worth pursuing.

They’re too small to pursue any single one as the market cap for a company, but presumably the fictional AI chip startup could pursue many of these smaller markets at once. It would be a long tail play, wouldn’t it?