The neural networks behind social media's sophisticated ad targeting could swallow the whole energy grid without ever getting to where their creators want them, scientists warn.
Neural networks should be used for more worthy ends than advertising if there is not enough energy in the world to train them to perfect.
The problem is that an infinite amount of energy can be used to train neural networks to target advertisements. It's like a black hole that swallows up all the energy you throw at it, which is not sustainable.
Abrahamsen said that the best conventional algorithms are only able to manage up to eight unknowns, and that the neural networks used by social networks to fine- tune their advertisements are no match for that.
Abrahamsen and his team concluded that even if all the energy on Earth was used to train a single neural network, it would never become perfect.
Things get slower and slower as we train neural networks, with a neural network only achieving 80 percent accuracy after a day, but only reaching 85 percent after a full day.
Artificial intelligence researchers are familiar with this appetite for power. Abrahamsen said the public doesn't know how harmful neural networks can be.
Abrahamsen said that they don't appreciate their contribution towards this enormous use of energy when compared to our awareness about the impacts of intercontinental flights or clothing purchases.
There is a growing body of research on Artificial Intelligence's massive carbon footprint, but it is mostly related to the other societal implications of these rapidly-improving technologies. The seedy energy of the Artificial Intelligence is also deserving of some sunlight.
Neural networks behind social media consume an infinite amount of energy.
Experts say that almost the entire internet could be generated by artificial intelligence.
Are you interested in supporting clean energy adoption? At UnderstandSolar.com, you can find out how much money you could save if you switched to solar power. Futurism.com may receive a small commission if you sign up through this link.