Right. But slapping boiling water around the burning plasma is kind of a rube goldberg usually. See LLNL's LIFE design for example [1]. Things like molten salt walls circulating through a steam turbine and all that.
There are other ideas too, but it's hard to beat a Rankine cycle.
You can't just slap boiling water around the burning plasma in a DT reactor, since you need almost all the neutrons to make more tritium. Water would absorb too many neutrons. The IFE designs use thick showers of liquid lithium or molten FLiBe for this reason.
I think we just haven't found the right fusion design.
If we use a reaction that primarily produces beta radiation or other high energy charged particle, sending it through a coil of wire would induce a voltage that we could extract as electric energy.
For that matter, appropriately located coils could be used to extract thermal energy from the plasma directly. The trick there is that we can't get much with the current tokamak and stellarator designs -- the thermal energy is too disordered to use a large coil and the plasma flow is not sufficiently confined to use small coils. There are almost certainly better configurations, but the electrohydrodynamics simulations are tricky. If we keep at it I'm sure we can find a stable configuration with fewer degrees of freedom.
I'm surprised too. I've looked into this before, and it's absolutely right - just not intuitive to me.
We do have radio-photo-voltaic devices, but they're so inefficient it's laughable. And we have RTG generators, which are only practical in limited situations, and again have a very low efficiency.
Well, there is hydro, wind and photovoltaic. And in the fusion field there are startups working on aneutronic fusion, which can generate power directly from charged particles. LPPFusion is one that seemed promising a few years ago, but unfortunately less so now.
There are other ideas too, but it's hard to beat a Rankine cycle.
[1] https://en.wikipedia.org/wiki/Laser_Inertial_Fusion_Energy