AI is the new cryptocurrency. It holds a lot of speculation and generates no profit. Of course investors will all kinds of nonsense to prop up their investment vehicle.

This raises lots of interesting questions. Off the top of my head, if this pans out:

- what would AI have to “deliver” in terms of capability to avoid collapse? And how much can be pawned off on the “people overestimate the short term value and underestimate the long term value” mantra, and for how long?

- what countries are most likely to come out ahead or less behind if the whole thing does collapse. Many European countries are investing heavily in AI but not quite going all in economically. How e.g. will France and Switzerland, who have big national investments, fare?

Not sure what I think about the theory of cryptocurrency being promoted as a tax dodge (for the wealthy) pending massive income tax hikes (on everyone else) to finance national AI spending. I'm not sure that passes the "why isn't this already happening, then?" test (crypto's being used as tax liability laundering, sure, but tax code restructuring doesn't seem to be in the offing) but it's interesting to consider.

I remain puzzled as to what it means to "win" at AI/have AI succeed in the macro, though--something that comes up in this article and others. Does that mean widespread automation of labor (at which point, whither goeth consumer spending)? Or like ... sci-fi bullshit a la Terminator and AI oppression? Or just a big-but-still-marginal productivity windfall in a bunch of industries (with concomitant unemployment/labor imbalance)?

Like, if (big if) the existential macroeconomic threat described here is real, the implication is that there will be an economy to save. In the future described by the AI boosters, will there be?