On 5 December 2023 John Kerry, the US climate envoy to COP28 in Dubai, launched an international ‘engagement plan’ to accelerate fusion research. Kerry’s plan involves cooperation between thirty-five countries on ‘R&D, supply chain and marketplace, regulation, workforce, education and engagement’. He presented fusion as a ‘critical piece of our energy future’ with the potential, if it can harness the atom to produce clean energy, to ‘revolutionize our world’.

My bet is that Kerry will be wrong and that fusion power will eventually join cold fusion and trans-warp conduits as neat but failed ideas. But fusion power appears to be a favoured technical fix for an intractable political problem around decarbonisation. We thus need to ask: what kinds of social functions are expectations about fusion performing? Focusing on whether fusion will work can mean we overlook the social management strategy involved in claiming that we should expect it to work. I suggest those social functions are fourfold.

Claims about fusion reify future innovation as a linear progression

Kerry’s marketing pitch for fusion power was that ‘we are edging ever closer to a fusion powered reality’. But consider two recent fusion episodes. Claims about the UK JET fusion reactor in 2021–22 and more generally about the French ITER fusion reactor, both of which use magnetic confinement to confine and heat plasma, were species of science communication hype. Thermal and electric inputs were under-counted and so net power gain was misrepresented. And claims about the US LLNL fusion reactor in 2022, which used inertial laser confinement to compress a fuel capsule, were species of manufactured ignorance. Both the total energy consumed to achieve ignition, and the military purposes of the experiment, were obfuscated.

Yet beyond the general inaccuracy of claims about fusion power, the specific accomplishments of fusion experiments—from demonstrating burning plasma to achieving ignition—have been amplified to frame technological promise as a linear inching forward. The perception of incrementalism is a social management exercise. Kerry’s future-oriented expectations are a technological promise akin to a contract with the future.

Fusion is just thirty years away?

Fusion is thirty years away now and will always be thirty years away, and that disappointment needs to be constantly massaged. Recently a few Japanese analysts performed a regression analysis of fusions researchers’ expectations, published between 1985 and 2022, about when fusion will be ready. They found that expectations are accelerating, or in other words that as we approach the present, fusion researchers expect fusion to arrive sooner—maybe in 17.8 years.

Yet our statistically precise analysts overlooked what is made obvious by the famous song ‘Tomorrow’ from the 1977 musical Annie: ‘Tomorrow, tomorrow, I love ya tomorrow. You’re always a day away’. Deferring fulfilment of a promise till tomorrow is less about pinging a precise date of success than it is about optimistically smiling through all the disappointments that continually litter yesterday. That is the real-world effect of the ever-present but-always tomorrow. As the economists Stephen Ziliak and Deidre McCloskey wrote in The Cult of Statistical Significance (2008), when trying to grasp ‘significance’ in human affairs, it is oomph (real-world effect), not precision, that matters most.

Expectations are performative

The outcomes of expectations about techno-futures do not hinge solely on whether gadgets come into being as imagined, as if technology determines social order, but ultimately upon organisational and cultural-political factors. In my field of Science and Technology Studies, we argue that technological expectations are therefore performative: they guide activities, provide pathways and political legitimation, attract investments and mobilise resources at various scales (national policy and regulatory design, intermediate public-private innovation networks and local research groups). Crucially, announced expectations about the future will often move from inflating the promise of research to intervening where failure might damage projects and reputations.

Kerry’s COP28 speech adopted a tone of humility regarding failed projections and likely engineering challenges, but also intervened in those past failures to situate them as actually inching towards a clean energy fusion future. He was there, he said, to ‘harness the power of fundamental physics and human ingenuity’, and that was the cue to pivot away from technology and towards people. Referring to visiting the so-called moonshot factory Google X in California, where he ‘listened to the scientists talking excitedly’ about fusion, Kerry also mentioned visiting ‘Commonwealth’.

What is Commonwealth?

Commonwealth Fusions Systems is an MIT startup, co-founded by Dennis Whyte (director of MIT’s Plasma Science and Fusion Centre) and designed to ‘leverage’ decades of research at MIT plus the ‘innovation and speed of the private sector’, which is ‘supported by the world’s leading investors in breakthrough energy technologies’. Commonwealth’s pitch is that it is using a new type of superconducting material to build a super-powerful magnet that will be better than ITER: still a tokamak, but stronger, smaller and cheaper. Though of course, this is the fusion power industry, so ‘Commonwealth’s magnet hasn’t been tested in a working reactor’.

The under-construction tokamak room at the Commonwealth Fusion Systems.
Cat Clifford, CNBC

Here we have our intervention, where failure might derail a project: Kerry, a former Senator for Massachusetts, spruiking for a public-private fusion startup in Massachusetts. Of course, Kerry was also introducing what is in effect ITER Part Deux, (another) thirty-five nations, this time led by the US in a policy direction—a ‘decadal vision’—announced by the US Office of Science and Technology Policy on 2 December 2023. Future expectations about fusion thus perform a rescue operation on the present, in this instance via a set of commitments to cooperate across several material, economic and political dimensions.

The quest for relevance

Expectations perform the social function of constantly laundering relevance. The ITER project is symbolic of the issue in this regard, because ITER was sold to politicians and to the public via misleading claims about net power gain, and specifically about a tenfold return on energy (Q=10): 500 MW out from 50 MW in. But ITER was only designed for net gain across the plasma, not the entire reactor, and so most estimates of net power gain ignore thermal and electric inputs. Fusion critics like Steven B. Krivit have called this the ‘grand illusion’ of fusion. ITER has now conceded that the net power of ITER will be zero and that its primary goal is to produce a burning plasma. The same deletions of inputs occur when conflating ignition with net power gain in inertial laser confinement experiments.

New future-oriented expectations have the social function of encouraging publics, politicians and investors with limited historical attention spans to not bother recalling failed projections. The British ZETA project (started in 1957) never worked as hoped and ceased operating in 1968, and the UK cancelled the Reversed Field Experiment (magnetic confinement) in 1981. The US scaled back its funding of fusion research from its heyday in the 1970s, with Reagan, Bush and Clinton iteratively cutting funding. One completed reactor, the LLNL Tandem Fusion Mirror Reactor, was never turned on, and eventually dismantled.

Returning to Massachusetts, Commonwealth’s Dennis Whyte had his fusion lab at MIT shut down in 2016, meaning that only computer simulations could be performed, not experiments. Whyte had claimed in 2013 to have designed a fusion reactor that could produce 250 MW of electricity. Commonwealth now claims that it has a great magnet to run a tokamak, that it is much better than ITER, and that by 2025 it will be ‘generating 10 times more energy than it consumes’.

It turns out the ITER myth travels from one fusion reactor to another (this one yet to be tested). Maybe that is the ultimate point about techno-expectations—that they can live fruitful lives without ever having to come true. What matters are not the claims themselves, but what others (hopefully) do with the claims. Fusion edgelords.

Nuclear After-Life: From tragedy to farce, the claims of a nuclear renaissance

Darrin Durant, Mar 2023

Non-hydro renewables have now overtaken nuclear power, with wind and solar alone reaching 10.2 per cent of global gross power generation in 2021.