The boom in artificial intelligence has already modified our understanding of technology and the world. However, developing and updating AI programs requires one loads of computing power. This relies heavily on servers in data centers, which incurs high costs when it comes to CO2 emissions and resource consumption.
One very special energy-intensive task is “training”where generative AI systems are exposed to massive amounts of knowledge so that they can improve their work.
The development of AI-based systems is blamed for this a rise of 48% on Google Greenhouse gas emissions over five years. This will make it harder for the tech giant to fulfill its goal of reaching net zero by 2030.
Some within the industry justify the extra energy required by AI by pointing to advantages the technology could have for environmental sustainability and climate protection. Improving the efficiency of solar and wind energy Predicting weather patterns, “smart” agriculture and more efficient, autonomous electric vehicles are amongst those rumored Benefits of AI for Earth.
With this in mind, technology corporations have been looking on renewable energies And Nuclear fission to power their data centers.
Nuclear fission is the form of nuclear energy that has been used worldwide for a long time. It releases energy by splitting a heavy chemical element into lighter ones. Division is one thing, but some in Silicon Valley feel it one other technology To close the gap, it can be essential: ​​nuclear fusion.
Unlike nuclear fission, nuclear fusion creates energy by combining two light elements right into a heavier one. But fusion energy is an unproven solution to the sustainability challenge of AI. And tech CEOs' enthusiasm for this technology as an AI powerhouse risks overlooking its potential advantages for the planet.
Beyond the standard
Google recently announced this signed a contract to purchase energy small nuclear reactors. This is a technology based on nuclear fission that makes it possible to generate useful amounts of electricity using much smaller devices than the large reactors of huge nuclear power plants. Google plans to make use of these small reactors to generate the electricity needed for the increasing use of AI.
This yr Microsoft announced an agreement with the corporate Constellation Energy, which could pave the way in which for the restart of a reactor on the Three Mile Island nuclear power plant in Pennsylvania, the positioning of the worst nuclear accident in US history.
However, nuclear energy produces long-lasting radioactive waste that have to be stored safely. Nuclear fuels similar to the element uranium (which have to be mined) are finite, so the technology will not be considered renewable. Renewable energy sources similar to solar and wind energy suffer from “intermittency,” meaning that they don’t produce energy consistently in any respect times of the day.
jamesonwu1972 / Shutterstock
These limitations have led some to contemplate nuclear fusion as an answer. This is most notable with Sam Altman of OpenAI shown particular interest In Helion energya fusion startup working on a comparatively novel technological design.
In theory, nuclear fusion represents a “holy grail” energy source since it produces a considerable amount of energy from small amounts of fuel without producing greenhouse gas emissions and producing comparatively little radioactive waste. Some types of fusion depend on a fuel called deuterium, a type of hydrogen that may be obtained from an abundant source: seawater.
In the eyes of its proponents like Altman, these properties make nuclear fusion well suited to fulfill the challenges of growing energy demand within the face of the climate crisis – and to fulfill the big demands of AI development.
However, while you look beneath the surface, the image will not be so rosy. Despite the hopes of their proponents, fusion technologies have done this don’t yet produce a sustainable net energy yield (more energy than is used to run the reactor), let alone produce energy at the dimensions required to fulfill the growing demands of AI. The merger would require many more technological developments before it will possibly fulfill its promise of adding electricity to the grid.
Wealthy and powerful people, just like the CEOs of giant tech corporations, can do that strongly influence how recent technologies are developed. For example, there are numerous different technological ways to perform nuclear fusion. But the precise path to fusion that is helpful for meeting AI's energy needs might not be the one which is right for meeting people's general energy needs.
Dil_Ranahunga/Shutterstock
The overvaluation of innovation
Innovators often assume that their work will result in ideal social outcomes. If nuclear fusion works on a big scale, it could make a helpful contribution to decarbonizing our energy supply because the world tries to tackle the climate crisis.
However, the humanitarian guarantees of each fusion and AI often appear to be sidelined in favor of scientific innovation and progress. When you take a look at those that put money into these technologies, the query actually arises as to who actually advantages from them.
Will investments in fusion for AI purposes enable its wider adoption as a clean technology to switch polluting fossil fuels? Or will a vision of the technology promoted by powerful tech corporations restrict its use for other purposes?
Sometimes it looks like innovation is the goal without considering the broader implications. This vision is harking back to Meta CEO Mark Zuckerberg's motto “Move fast and break things”where short-term losses are accepted with a view to pursue a vision of the long run that may later justify the means.
