I guess it becomes okay when the companies guzzling the energy are some of the biggest tech employers in the world, buttering your bread in some way.
Whether the cost/benefit works out in the case of AI is another question.
On one hand, the cost of compute per token has gone down a lot, and will continue to go down, because that's exactly the economic incentives at play. We had a little short-term nonsense where "the bigger the better" was all the rage, but inference was never this way, and now training is also pushing in this direction.
But on the other hand, less compute per token means it can be more broadly deployed. And so there is likely more energy use, not less, in the long run.
Economic chaos is one thing, Skynet is another. And then people with the power are stupid and evil. Joy.
Companies like Apple and Google are both building data centers and trying to make on-device AI a thing. Unfortunately, they also keep inventing new, more expensive algorithms.
It’s at least plausible that most LLM use will become cheap enough to run on battery-limited devices like laptops and phones, though it’s not what most people are betting on.
[1] https://epoch.ai/data-insights/llm-inference-price-trends
VR is back in its niche.
3DTV .. I've not seen that marketed for quite a while now. Things that are fads will die out eventually.
Crypto meanwhile is in an odd space: everyone knows blockchains are too much of a hassle and that the volatility is too high, so they're using centralized exchanges and "stablecoins" (shadow dollars). There's still a huge amount of money there but not quite as much as the stadium ads / FTX peak.
Its probably coming back in a few years via lightfield displays, though for some reason Google seems to think the killer app for those is videoconferencing.