Serious Question About Latvian Gambit

I recall an article either in Chess Life or New in Chess magazine where Caruana declared with unassailable authority that the Evans Gambit was dead and forever refuted. Now recently AlphaZero played it. I am doubting anyone could refute the gambit against AlphaZero except maybe a stronger program. So the common judgments of strong players are not so absolute in light of stronger entities. Fine. Can we extend this to a revaluation of the Latvian Gambit? I haven’t seen AlphaZero play this gambit, but I suspect that it would demonstrate very playable lines. Has anyone seen such a demonstration or thought of this? I know the best players clearly conclude that this Latvian opening is bad. I’m just wondering if that judgment might merely be the consensus of mortals, whereas computer AI tutelage would teach otherwise. It’s surely strange to think of a revaluation here given the level of certainty with which it’s deemed refuted and dead - but that seems to be what happened to the Evans Gambit.

I'm pretty sure the latvian gambit is dead at high level. I'm not very knowledgeable about the opening, though. You'll have to hope someone else who knows more about the subject comes around.

Yes I think it’s probably dead, but I’m wondering if a super strong AI could revive it. An AI could probably beat a grandmaster regularly using even the BongCloud.. so I start to wonder whether chess is richer than previously thought, even by grandmasters. If an AI can revive any dubious opening then what does that say? Is chess theory just a cultural construct, which like science, must always stand open to revaluation and higher order truth? Is everything beyond forced mates and below say 8 or 10.0 advantage open to revaluation?