Move over fission, nuclear fusion is coming, with the help of AI.
Reading the news of AI controlling a Nuclear Fusion reaction blew me away, almost in the same way that the experiment itself had all but done. For not being a subject matter expert, it is absolutely mind-boggling to take even the slightest interest into the interconnection between these two areas of science. The ultimate goal of the research is to create Nuclear Fusion.
What is Nuclear Fusion? Firstly distinguish it from Nuclear Fission, which is what we are familiar with from Chernobyl to Fukushima as the notion of nuclear energy.
In the simplest terms nuclear fission involves breaking apart Uranium and in doing so releases energy. This is nuclear energy as we know it. The breakdown of nuclear particles generates energy, a lot of energy, and unfortunately also means it is unstable and leaves lingering trace elements sometimes very harmful and longlasting.
Nuclear Fusion does kind of the opposite to Nuclear Fission. Rather than separating the atomic matter, it fuses them together. This is not an easy task. In fact it requires a lot of energy as an input to create the energy output. However, if the process can be refined and perfected, the output (at least until now it was theoretical) will outweight the input.
Another helpful graphical illustration shows Deuterium and Tritium, both elements in the form of isotopes of hydrogen are fused together, with the resulting chemical reaction in the form of helium and energy.
Credit: www.Byius.com
The benefits of Nuclear Fusion over Nuclear Fission include:
Recommended by LinkedIn
Currently the problem remains being able to generate the power to casue the fusion reaction and being able to that in an efficient and stable manner. This is where AI comes in.
Scientists at the Joint European Torus (JET) near Oxford, UK, announced on 9 February that they had generated the highest sustained energy pulse ever created by fusing together atoms, more than doubling their own record from experiments performed in 1997.
In a joint effort together with the Swiss Plasma Center (SPC) and artificial intelligence (AI) research company DeepMind, scientists used a deep reinforcement learning (RL) system to study the nuances of plasma behavior and control inside a fusion tokamak – a donut-shaped device that uses a series of magnetic coils placed around the reactor to control and manipulate the plasma inside it.
The Joint European Torus tokamak reactor near Oxford, UK, is a test bed for the world’s largest fusion experiment — ITER in France.Credit: Christopher Roux (CEA-IRFM)/EUROfusion (CC BY 4.0)
In order to sustain nuclear fusion reactions it is necessary to raise the plasma and keep it stable at hundreds of millions of degrees Celsius, which for reference is hotter than even the core of the Sun. AI assists but maintaining the temperate of the plasma by managing coils to keep the heat rising and constant. AI is playing a critical role in assisting bringing the requisite conditions for the fusion reaction to be undertaken.
The research journey is far from over and it will be at least 2025 before we see the private sector finish making the first reactors, but this kind of technology is a good example of all those incremental steps and hard work that has occurred for so long, suddenly bursting forward into the public domain.
I am still learning about this area of energy and the science is fascinating, the hope is that this energy can be developed on smaller scales and in a decentralised manner. That really is the key. The problems we have in the world today, perhaps ironic being nuclear fusion and not fission (given the existing Ukraine-Russia crisis), are all problems stemming from the geopolitical tensions of energy.
We rely on energy sources to be processed using massive amounts of industry. If oil or gas, solar or coal, could be produced by individuals, on their roofs without the dependence on government, we would have much more freedom in our society. Whether nuclear fusion can be an open-source technology - like the internet - only time will tell.