When researchers were saying, "Fusion is twenty years away" (not ten), it was also during a boom in fusion power funding.

Then the funding collapsed.

We still haven't spent the money that was expected to be spent in those "twenty years", so I think the jury is still out on this one. Certainly, there has not been deliberate fraud in the field, so I think the phrase "huckster" is unjustified.

I am still pretty skeptical about fusion power, but at least we know fusion itself works - just look up on a sunny day. AI does not have such a canonical example.