I feel your argument relies on assuming that being an optimist or pessimist means believing 100% or 0%, whereas I'd claim it's instead more just having a relative leaning in a direction. Say after inspecting some rusty old engines a pessimist predicts 1/10 will still function and an optimist predicts 4/10 will function. If the engines do better than expected and 3/10 function, the optimist was closer to the truth despite most not working.
Similarly, being optimistic doesn't mean you have to believe every single early-stage invention will work out no matter how unpromising - I've been enthusiastic about deep learning for the past decade (for its successes in language translation, audio transcription, material/product defect detection, weather forecasting/early warning systems, OCR, spam filtering, protein folding, tumor segmentation, spam filtering, drug discovery and interaction prediction, etc.) but never saw the appeal of NFTs.
Additionally worth considering that the cost of trying something is often lower than the reward of it working out. Even if you were wrong 80% of the time about where to dig for gold, that 20% may well be worth it; reducing merely the frequency of errors is often not logically correct. It's useful in a society to have people believe in and push forward certain inventions and lines of research even if most do not work out.
I think xvector's point is about people rehashing the same denunciations that failed to matter for previous successful technologies - the idea that something is useless because it's not (or perhaps will never be) 100.0% accurate, or the "Until it can do dishes, home computer remains of little value to families"[0] which I've seen pretty much ad verbatim for AI many times (extra silly now that we have dishwashers).
Given in real life things have generally improved (standard of living, etc.), I think it has typically been more correct to be optimistic, and hopefully will be into the future.
[0]: https://pessimistsarchive.org/clippings/34991885.jpg