This is the heart of the Z-machine ("zee machine"). Looking and sounding like something from science fiction, this is today's science fact. It's one of the research projects that might - just might, with enough funding and enough will - solve the world's energy problems. Or, alternatively, it might not. But if it doesn't there are other fusion projects waiting in the wings to secure what is probably humanity's greatest prize: freedom from our dependence on fossil fuels. From a disappointingly slow start this episode of Horizon grew into one of the most fascinating glimpses into one possible future that I've ever seen.

I know there are those for whom Brian Cox is just about the best hard science presenter on the telly. Unfortunately, I'm not one of them. Whether it's his permanently-affixed dopey grin, his soporific voice, or the way he approaches his subject from somewhere out beyond the Lesser Magellanic Cloud and arrives, via Saturn and that small colourful village somewhere just outside Wales, to discover that he was only popping next door to borrow a cup of sugar, I'm not sure. Maybe it's all of the above.

He is enthusiastic though. And knowledgeable. It's just a shame he talks to the rest of us as if we're all at the level of amoeba. Veeerrryyyy sllooowwwllyyy, in Simple Language (with lots of capitals).

So it was that in a 50-minute programme supposedly about fusion and whether we can do it on Earth, we spent the first 35 minutes going around the houses of the Sun and how it works, Einstein's Theory, Eddington's Realisation, how the world is dependent on fossil fuels, the energy density of oil, the atom bomb and all sorts of stuff designed to make the last 15 minutes appeal to an audience that had come from a standing start and hadn't even the foggiest idea of what fusion meant, except that it might have had something to do with sticking something to something else. That'll be the audience that wasn't watching in the first place then? Still, at least he probably hit some government target about making science more accessible.

The figures were impressive though, and Our Brian spewed them out like hot jets of gas coming out of the Sun. 4 million tons of the Sun's mass is burned off every second! Wow! There's the same amount of power in a dollar bill as in an atomic bomb! Wow! And so on.

In among all this house circumnavigation there was a rather elegant thought experiment that really appealed to the geek in me. **WARNING** Geeky stats paragraph alert. Coxy was discussing average energy consumption and whether it was possible to generate enough power for everyone using alternative (non-fossil) fuels. They decided 5kW was a good average power consumption figure per person (Americans use considerably more than this, Europeans less, etc) to give everyone a comfortable standard of living. With 6billion of us on the planet, this means we need generating capacity of 30 terawatts (30TW). They set themselves a target to have this in place by 2035, and divvied up the power between all the available forms of alternative energy. Giving one-fifth of the requirement to nuclear power means that we, as a world, would need to build two-and-a-half nuclear power stations every WEEK for the next 25 years. It's worse for solar power. We'd need to deploy 250 square metres of solar panels every SECOND for the entire 25 years to meet the demand (I think they only gave solar something like 5% of the need). And that's before they factor in population growth.

**END of geekiness

Basically, it's simply not credible to generate enough power to give everyone on the planet a Western lifestyle in 25 years time using existing green fuels. Fusion is, pretty much, our only hope. So - FINALLY! - can we do it? And this is where the programme started to really get interesting.

With, first off, a visit to the z-machine, that takes a day to set up, causes a minor earthquake when it's fired, destroys several hundred tiny wires and generates enough power for the entire United States for a fraction of a second. It causes a "flashover" which can be seen here. Now, if they could only cut the time to cycle the reactor from a day to a few seconds, it could be a real generator.

Over in Korea, they're trying to do the same thing with supercooled magnets in a machine called the K-Star. And in a different part of the US, at the NIF (National Ignition Facility) they're focussing a hundred or so lasers onto a very small ball of material - about the size of a pinhead - which makes it collapse and start a very brief fusion reaction.

So the answer to the eponymous question is yes, briefly. We - that is, scientists - have done it. If you'd asked "can we generate power from the star we've made on Earth?" the answer would have to be no, for now. Estimates of exactly when we will be able to do that vary from 2022 to 2050. If they're anything like the estimates that used to be dished out on Tomorrow's World I wouldn't wait around to see any fusion power in your neighbourhood until 2100 at the earliest, but that's not taking into account the increasing social, political and climatological pressures that will bear down on the fusion scientists in the years to come.