How Much AI Do We Really Need?
Just when we thought we might be able to squeeze through the minefield of climate change tipping points, along comes Muskrat and his cadre of Tech Bros eager to get their greedy little paws on Artificial Intelligence so they can make their next billion bucks.
They want to use AI to sell things like more tedious undergraduate papers, fake news stories like the recent whopper that Luigi Mangione had killed himself attributed to the BBC, be our Chat-Box friend, lover or confidant or simply to annihilate humanity through an unstoppable pandemic or world war.
The problem is, AI will need over 85 terawatt hours of electricity by 2027 to mine our personal data so it can fleece us out of our hard-earned cash.
So, Musk and his boys are scrambling to get subsidies for nuclear energy, both fission that continues to be super dangerous, and fusion that will need a lot more time, money and research to make work.
One thing the Tech Bros have to figure out is how to use fission energy or fracked natural gas to jump start their fusion nuclear reactors.
All this, to simply boil some water to make steam, to make a turbine spin, to make some electricity.
Instead of such a Rube Goldbergian solution, wouldn’t it make more sense to simply use wind to spin a turbine, or use a photovoltaic cell that seemingly uses magic to convert sunshine into electricity.
The problem is, that we need energy to have a good life and preserve our civilizations, without destroying our planet.
But do we really need AI?
Before we build a lot more expensive, dangerous, polluting sources of energy shouldn’t we decide just how much AI we really need, or if we really need it at all?