Two weeks ago, we talked about how artificial intelligence (AI), crypto mining, and clean energy manufacturing are driving the Fourth Industrial Revolution (simply the 4Rs), along with data and connectivity, analytics, and human-machine interaction. , improvements in robotics. Unfortunately, these long-term megatrends are pushing the U.S. power grid to its limits.
Sreedhar Sistu, Vice President of Artificial Intelligence, said: schneider electric (OTCPK:SBGSF), excluding China, AI could account for 4.3 GW of global electricity demand, increasing almost fivefold by 2028. Another analysis predicts that demand from AI will increase exponentially, increasing at least 10 times between 2023 and 2026.
AI tasks typically require more powerful hardware than traditional computing tasks. Meanwhile, Bitcoin mining shows no signs of slowing down, with mining rates currently reaching 565 exahashes per second (EH/s), a five-fold increase compared to three years ago.
Bitcoin mining consumes 148.63 TWh of electricity per year and emits 82.9 million tons of CO2 per year. This is comparable to Malaysia's electricity consumption. And the demand for data centers is not helping matters at all. Data center storage capacity is expected to grow from 10.1 zettabytes (ZB) in 2023 to 21.0 ZB in 2027, at a favorable CAGR of 18.5%.
An analysis by Boston Consulting Group predicts that data center power consumption will triple by 2030, enough to power 40 million U.S. homes.
The situation is already getting out of hand, with U.S. electricity demand beginning to rise for the first time in 15 years. ”We as a country are short on energy;“Michael Khoo, director of the climate disinformation program at Friends of the Earth and co-author of the report on AI and climate, told CNN.
To be fair, AI has been touted as one of the key technologies that will help tackle climate change. This innovative technology is already being used to track pollution, predict the weather, monitor melting ice, map deforestation, and more. A recent report commissioned by Google and published by Boston Consulting Group claims that AI could help reduce global warming pollution by up to 10%.
Unfortunately, past trends in technological progress suggest that the disadvantages of AI will very likely outweigh the advantages as far as electricity demand is concerned.
”Even with increased efficiency, the energy consumption of cryptocurrency mining never decreased. Making a particular good or service more efficient increases demand.” says Alex de Vries, a data scientist and researcher at Vrije Universiteit Amsterdam.
At this point, almost everyone agrees that we do not have the capacity to develop renewable energy power plants fast enough to meet this rapidly increasing demand for electricity. So other than just saying let's build more natural gas and fossil fuel power plants, what other options do we have?
Enter nuclear fusion, long considered by scientists to be the holy grail of clean, nearly limitless energy. Sam Altman, head of ChatGPT creator OpenAI, says fusion is the ultimate solution to AI's energy puzzle“There is no way to get there without a breakthrough, we need fusion.” Altman said in an interview in January. Altman reiterated this view a few weeks ago when podcaster and computer scientist Rex Fridman asked him about the AI energy conundrum.
blue sky thoughts
Unfortunately, Altman's proposal is likely to be another case of overly optimistic, bolt-from-the-blue thinking, getting no closer to building a commercial fusion reactor than harvesting energy from black holes. There is a possibility.
For decades, nuclear fusion has been considered the “holy grail” of clean energy. If we could harness that power, we could create an endless amount of clean, sustainable energy. It powers stars, and theory suggests it could be successfully applied to nuclear reactors without the risk of catastrophic meltdown disasters.
Scientists have been researching practical fusion reactors since the 1950s and are hopeful that a breakthrough is just around the corner. Unfortunately, practical fusion power plants may be decades or even centuries away, and there are jokes going around about milestones being missed multiple times.
To be fair again, there have been some promising signs about the possibilities here. Last year, a fusion reactor in California produced 3.15 megajoules of energy with just 2.05 megajoules of energy input. This is a rare example of a fusion experiment producing more energy than it consumed. Most fusion experiments are energy negative, taking in more energy than they produce, making them useless as a form of electricity generation. Despite growing hopes that nuclear fusion could soon play a role in mitigating climate change by providing large amounts of clean electricity for energy-intensive technologies like AI, the world teeth”We are still far from commercial convergence and cannot currently solve the climate crisis.'' said Aneeka Khan, a fusion researcher at the University of Manchester. guardian Shortly after the initial December breakthrough.
You don't have to look far to see the full reality.
For decades, 35 countries have collaborated on the largest and most ambitious scientific experiment ever conceived: the International Thermonuclear Experimental Reactor (ITER), the largest fusion power generator in history. I did. ITER plans to generate plasma 10 times hotter than the sun's core, producing net energy for seconds at a time. As is usual with many nuclear power projects, ITER is already facing huge cost overruns, calling its future viability into question. W
When the ITER project officially began operations in 2006, international partners agreed to finance an estimated 5 billion euros ($6.3 billion at the time) for the 10-year plan, with the reactor coming online in 2016. charles sifThe director of New York University's Arthur L. Carter Institute for Journalism sued ITER, citing a lack of transparency about costs and persistent delays. He said the latest official cost estimate for the project currently stands at more than 20 billion euros ($22 billion), and the project is far from its primary objective. Worse, none of ITER's key stakeholders, including the U.S. Department of Energy, have given any concrete information on whether the team will be able to overcome the technical challenges or the additional delays, much less the estimated additional costs. Failed to provide answers.
Source: Scientific American
Notre Dame Cathedral took a century to complete, Seife said, but it was finally used for its intended purpose less than a generation after construction began. But the same cannot be said for ITER, which he concludes will look less and less like a cathedral and more like a mausoleum.
Written by Alex Kimani, Oilprice.com
Other top articles on Oilprice.com: