Respectfully, I disagree. We’ve entered an AI boom, and right now, the star of the show is in a bit of a gangly awkward teenage phase. But already, these large data models are eating up mountains of energy. We’ll certainly make the technology more energy efficient, but we’re also going to rely on it more and more as it gets better. Any efficiency gains will be eaten up by AI models many times more complex and numerous than what we have now.
As climate change warms the globe, we’re all going to be running our air conditioning more, and nowhere will that be more true than the server centers where we centralize AI. To combat climate change, we may figure out ways of stripping carbon from the air and this will require energy too.
Solar is good. It’s meeting much of our need. Wind and hydroelectric fill gaps when solar isn’t enough. We have some battery infrastructure for night time and we’ll get better at that too. But there will come a point where we reach saturation of available land space.
If we can supplement our energy supply with a technology that requires a relatively small footprint (when it comes to powering a Metropolitan area), can theoretically produce a ton of power, requires resources that are plentiful on Earth like deuterium, and doesn’t produce a toxic byproduct, I think we should do everything in our power to make this technology feasible. But I can certainly agree that we should try to get our needs completely met with other renewables in the meantime.
