Oil Companies Are Using AI Bulls**t To Suppress Renewable Energy
They aren't even being subtle about it.
You may have heard of Sultan al-Jaber, as he was the oil billionaire who led last year’s COP28 climate summit. Seems like a bit of a conflict of interest, right? Well, he recently told the Financial Times that the immense rise of AI is giving the world’s largest oil companies a major incentive to increase their investments in renewable energy. On the surface, this sounds like a good thing. But if you even take a second to peer beneath the surface, you will find stinks of oil industry greed and manipulation and the cost of our planet’s health.
So, what reason did al-Jaber give for this view? Well, he cited how Shell and BP have massively pulled back from their renewable push over the past two years. But, are now interested in supplying renewable energy to AI companies (specifically generative AI).Â
You see, AI companies require vast amounts of energy to train and run their giant Large Language Model (LLM) AIs, such as Chat-GPT. These same companies have also pledged to become net-zero in just a few years, as the tech industry is seen as the easiest to decarbonise, and these companies don’t want to be restricted by climate legislation. But AI is ruining these plans. Google’s emissions have increased by 48% over the past five years thanks to their AI push. So, big tech is looking to buy vast amounts of low-carbon renewable energy and even nuclear power.Â
In other words, not only is the AI boom dramatically increasing our overall energy demand, but the increased demand is for low-carbon renewable energy. Not wanting to miss out on an oportunity, oil companies want to be the ones supplying this energy and profiting from it.
That sounds good, right? Finally, capitalistic, free-market forces are driving investment into energy sources that don’t destroy the planet.
Well, no.
Al Jaber also recently said that renewables and nuclear simply can’t meet the surging energy demand of AI. As such, he insists that natural gas will play a key roll.
Bingo, that is the issue. Quite simply, the Sultan is talking out of his arse.
Firstly, 100% renewable energy grids are absolutely possible.Â
For example, a recent study found that the UK can supply 90% of its energy needs from wind and solar without raising energy costs. But 15% of the UK’s energy already comes from nuclear power. As such, the UK is a hop and a skip away from being as low carbon as possible.
Another novel study found that the entire continental US can run on 100% wind and solar energy without the need for any grid-level storage. The idea is that if enough solar panels and wind turbines are spread around the country, at least one location will generate enough energy for the entire country, negating the need for expensive battery storage. You might think using such a vast volume of renewables would be too costly and inefficient with land usage, but you’d be wrong. Using real-life weather data, this study found that such an energy grid with enough renewables to have no blackouts would only take up 0.84% of US land, substantially less than the 1.3% that is currently taken up by the fossil fuel industry. This study also found that per capita energy expenses would be 63% cheaper than the US’s current energy grid!
The narrative that these oil companies are pushing is that the extra energy demand of the AI industry will make these 100% renewable or ultra low carbon energy grids impossible, as they simply can’t supply enough energy. But that isn’t true.
The most extreme predictions suggest the generative AI industry will use around 1,000 TWh of energy per year by 2030. Meanwhile, the annual global energy consumption is predicted to increase to 36,000 TWh by 2030. As such, AI will only account for a 2.7% increase in total energy demand. This is a minuscule increase in the grand scheme of things and can be easily accommodated within ultra-low carbon energy grids.
What’s more, we aren’t even sure if AI will require such a huge impact. As OpenAI has shown, current AI models simply aren’t profitable as they are too expensive to develop and maintain (read more here). We are also hitting diminishing returns with AI training (read more here), and as such, for AI to get even a little better, it will require exponentially more energy. But with its lack of profitability, can these technology companies justify such an expense? The issue is further compounded by the fact that predictions of AI’s impact on productivity and GDP are getting more and more pessimistic. Some even suggest it can only increase productivity by 0.5% and increase GDP by 0.9% cumulatively over the next decade. As such, many don’t think AI will have the wide adoption some have predicted it will have.
With this final piece of the puzzle, we can finally understand what al-Jaber is doing.
al-Jaber is doing a very bad job of covering up what he is actually trying to do here. He is a huge supporter of AI, and saying that AI is incentivising more renewable energy gives the industry a huge PR boost. However, in reality, he and his oil friends are trying to shift renewable investment into AI, an industry they know likely won’t live up to the hype, and not into the public energy grids. This will enable their main business of killing the planet and selling us fossil fuels to continue for longer.
Meanwhile, in reality, if we actually invested as much as we should be in renewables, rather than paying these oil billionaires yet more trillions of dollars each year through tax breaks and subsidies, we could build 100% renewable energy grids that can power us and the AI boom easily, without raising energy costs, and without killing the one planet, we can call home.
Thanks for reading! Content like this doesn’t happen without your support. So, if you want to see more like this, don’t forget to Subscribe and help get the word out by hitting the share button below.
Sources: FT, Recharge, The Guardian, Fortune, Will Lockett, Will Lockett, CDO Trends, BBC, The Earth & I, Statista, ComputingÂ