When it becomes cost effective relative to other cheap options. This might take a while. We haven't yet reached the viable prototype stage with a prototype that

When it becomes cost effective relative to other cheap options. This might take a while. We haven't yet reached the viable prototype stage with a prototype that 1) generates more power than it consumes and 2) is practical and cheap enough to operate for extended periods of time. That wouldn't be good enough though because it would likely still be far more expensive than other things in the market. So we need a third condition: 3) is significantly cheaper than anything else. That's a tough one. It might never happen. We've had proofs of concept that prove it is is possible to generate more power with fusion than you put in. That happened fairly recently. So, we sort of have the first of the 3 checked off. The rest is going to take lots of time. So, fusion may happen; and I actually believe it will. But it will likely be costly for the foreseeable future. And when I say costly, I really mean prohibitively expensive for all but the most desperate of situations where literally nothing else can do the job. From there to actually being cheaper is going to be a long journey. 2100 is about the earliest I can see this working. But frankly, I think its optimistic. As the article points out. Renewables work now and they work cheaply. And they are already scaling faster than anything else in the energy sector. We'll have decades more to do better than we are already doing. Renewable prices dropping by one or two orders of magnitude would shut down most of our legacy energy providers. That's what fusion is up against. It's a race to the bottom. Not a great context for what looks like it's going to be a highly capital intensive thing with essentially no ROI for the foreseeable future.