> [...] but nobody will invest dollars backing a SaaS offering that could easily be replicated, or that could be made redundant tomorrow.

Matt Levine wrote in his newsletter Money Stuff of some investment fund that has their employees vibecode replacements for software of potential investment targets. I guess the theory is exactly what you say: if the internal employees can replace the target's software in a few hours, that's a big signal on whether to invest or not.

(I wouldn't quite say you shouldn't invest at all; but you have to argue that the moat is in eg the sales process or the existing customer base or network effects etc. Even before AI, people famously build Twitter clones over the weekend for fun.)