AI is supposed to be the future. But it's being powered by the past.
Artificial Intelligence is generating tons of buzz. It can do everything from helping students write an essay to helping doctors diagnose diseases. But there’s another “buzz” AI is generating and that buzz you hear is coming from restarting coal and natural gas power plants, and building new gas plants, to feed the energy greed of the data centers. And in some cases, companies are extending the life of existing plants -- including 3 Mile Island! -- and planning new nukes.
Major tech companies including Google, Facebook, Microsoft, and Amazon are embracing AI as the best of what is to come. And they plan on profiting handsomely.
But all that profit comes at a deep, deep cost.
- The power being generated is polluting communities near power plants and the data centers. Many of these power plants are in low-income communities leading to more negative health effects, including air and noise pollution, in places that are already polluted.
- Instead of decreasing climate pollution as most of the companies pledged to do, their emissions are increasing. The World Benchmarking Alliance (WBA) and the International Telecommunication Union (ITU), a UN agency for digital technologies, report that the operational emissions of the four most prominent AI-focused technology companies increased by an average of 150% from 2020 to 2023.
- AI’s demand for energy is driving up electric rates across the country. Rates could surge by 20% this summer and there could be power shortages as utilities struggle to meet increased demand from data centers.
- AI has a growing hunger for electricity. By 2028, the researchers estimate, the power going to AI-specific purposes will rise to between 165 and 326 terawatt-hours per year. That’s more than all electricity currently used by US data centers for all purposes; it’s enough to power 22% of US households each year. That could generate the same emissions as driving over 300 billion miles—over 1,600 round trips to the sun from Earth.
- AI has a great thirst for water too. In some places, local residents are reporting that their taps are running dry from data centers taking all the water. And this thirst is only projected to increase.
- Besides the pollution and climate issues AI has many other problems as well including but not limited to copyright and privacy issues. AI has been shown to have bias and discrimination issues as well.
When it comes to the costs to society from all the pollution, there are solutions that can mitigate them. Solar, wind and battery storage are the cheapest sources of energy, and they don’t produce local pollution or climate emissions.
And these Big Tech companies know it. Google, Facebook, Microsoft, and Amazon already use renewable energy for much of their needs. But more must be done to ramp up renewable energy purchases and improve the efficiency of AI to reduce energy needs. And if these companies don’t take the lead, AI energy demand will only take us back to the days of more pollution and to a future where climate change is significantly worse.
