Related Post
An energy ‘breakthrough’ will be needed if AI is to fulfill its potential, says OpenAI boss Sam Altman.
The German News Agency quoted Altman as saying earlier in January: The energy needs of the artificial intelligence industry could depend on nuclear fusion, a technology that humans have not been able to develop, but which, if achieved, could greatly exceed the capabilities of… Current energy production.
“This motivates us to invest more in fusion technology,” Altman said at the World Economic Forum in Davos.
Scientists and engineers have been trying for decades to discover nuclear fusion power generation that partly mimics how stars like the Sun power themselves, but without the breakthrough that would make mass production viable.
Altman’s comments came after the publication of projections showing that artificial intelligence data centers will need the same amount of electricity as a medium-sized country by 2027.
These estimates were published in 2023, less than a year after OpenAI released its chatbot GBT for public use. Among the countries that compared estimates of the size of their electricity consumption is the Netherlands, which has a gross domestic product of about one trillion dollars, and is one of the richest countries in the world.
The energy consumption of data centers – which support billions of cloud computing, live streaming and social media users in the world – was controversial even before the recent expansion of so-called “generative” artificial intelligence, due to the increasing pressure it exerts on energy networks, and its impact on transformation. On consumer prices.
In Ireland, where many tech giants have European regional headquarters, electricity needs for data centers rose from 5% of available national supply in 2015 to 18% in 2022. Electricity supplies and prices have come under further pressure due to the Russian invasion of Ukraine in 2022, which It prompted Europe to make efforts to stop its dependence on Moscow’s gas exports.
