Until now, even the most optimistic fusion power advocates have said the technology won’t be putting electricity on the grid for at least another decade. Microsoft and its current court favourite, Sam Altman, say that is about change. Altman has invested $375 million in Helion Power, a start-up betting the house on an unproven device called a plasma accelerator, and the WSJ says Microsoft has signed a deal to buy at least 50 gigawatts of power from Helion within a year of its self-imposed deadline for success of 2028. Unlike most fusion power designs, the accelerator draws power directly from the magnetic chamber hosting the fusion reaction rather than converting heat to steam to drive an old-fashioned turbine. But no sustained fusion reaction has ever been achieved in an accelerator, and the direct-to-grid plan is no more than a plan. Has Microsoft put too much faith in Altman just because he’s produced a winner in OpenAI?