I got a nice mention in the Observer yesterday, in a review of the George Lamb Show on 6 Music. When you do these things, you forget someone might be listening, let alone that someone might be reviewing it.

Anyway, gratifyingly, Miranda Sawyer says George has “interesting guests”, I am the “regular clever bloke”, and I was “fascinating” when talking about the Large Hadron Collider:

“According to Brooks, one of the reasons why the LHC might keep going wrong is because time travel is too problematic to be allowed and someone/thing from the future is coming back to stop the collider working and save us all from broaching the space-time continuum. Yeah!”

Just wanted to throw in a note of clarification. Before any of the world’s physicists turn on me (again) I wasn’t quite saying that. I was saying that a couple of quite respectable physicists have suggested it. It’s appealing, especially when a couple of Russian mathematicians have suggested that the LHC could rip holes in spacetime and allow time travellers from the future to visit. But if that seems far-fetched (fun, though, huh?!), the new idea is just plain crazy. They seem to have forgotten Occam’s razor: the simplest explanation is usually the right one. And when you know that a couple of solder joints are the cause of the failure, there’s no need to invoke interference from the future.

The LHC is due to start up again in mid-November. If it fails for some inexplicable reason, then maybe we can invoke Stephen Hawking’s Time Cops. But until then…

The Guardian took me by surprise yesterday and published an op-ed piece I’d written for Monday’s paper (i.e. two days early). Which is great, because far more people read the Saturday Guardian.

It’s about the limits of science, and puts together two pieces of research. One is a new paper about the black holes that the LHC might produce (thanks to the physics arXiv blog for a neat summary, by the way, and you can get to the paper via that link). The other is something I came across when I went to a conference on Global Catastrophic Risk in Oxford last July. A group of researchers there have shown that any scientific argument has a finite chance of containing a fatal flaw, and so should really be presented only when accompanied by a probability statement about how reliable the reasoning or results are likely to be. Presumably, the same goes for this paper, too…anyway, Mark Buchanan has written about it for New Scientist at greater length, and that page contains links to the paper.

The point is, the two come together to show that there are limits to what science can tell us – and that the safety of the LHC is no exception. "What science can’t tell us" is the subject of a debate I’m doing with Rupert Sheldrake at London’s Institute of Contemporary Arts this coming Monday. Which is why I wrote the piece for Monday’s paper. But, as I said, I’m not complaining…

The piece about Global Catastrophic Risks I did for New Scientist is here, by the way. The end of the world is probably not nigh, as it turns out.  

I was unbelievably cheered by the news that the Large Hadron Collider has a helium leak. Not because I’m a nasty person, but because I’ve experienced so many of these myself. It’s just nice to know even the big guns can’t get it right all the time.

During my PhD research (that's NOT me in the picture), I had to operate a cryostat that cooled samples of superconducting metal down to 0.3 Kelvin (-272.7 celsius). To do this, you first cool to 77K with liquid nitrogen, which is fairly easy, and then you get down to 4.2K with liquid helium. (For me, the final stage was to use helium-3 to get down to 0.3K, but the LHC doesn’t need to do that).

SO MANY TIMES, I’d spend a couple of days getting my cryostat down to liquid helium temperatures only to find that an electrical connection somewhere in my experiment couldn’t handle the contraction due to temperature, and dropped off, making the whole rigmarole of cooling a colossal waste of time. Then I’d have to let the thing warm up again – another day or two. Only then could I get the experiment out, do the repairs and start the whole process of evacuating and cooling the apparatus again.

At the LHC, it seems they had the same problem on a bigger scale. The spokesperson said, “the most likely cause seems to be a faulty electrical connection between two of the magnets which probably melted, leading to a mechanical failure.”

Unfortunately for the LHC people, it takes two months, not a few days, to warm up, repair, and then recool their experiment. But the point is the same: doing science can be difficult, frustrating and – dare I say it – sometimes boring. Many people who read the news reports about the LHC’s problems will be aghast that they have lost two months, and consider this a major failing. It’s not, it’s just the way it is. The exquisite knowledge we gain from modern physics experiments comes at a high price: it’s cutting edge stuff, and thus expensive and, above all, time-consuming. But ultimately worth it, if the results of past experiments are anything to go by. It’s amazing how much we know about the universe from past accelerator experiments. The only particle of the standard model of physics we haven't yet seen is the Higgs boson, and that might not be hidden for long.

So, for the record, I’m still excited by the LHC. And now I’m empathetic too…
 

Apparently, scientists are getting death threats. Does anyone really believe the Large Hadron Collider will destroy the planet in a few hours’ time? Surely they don’t. It’s not like anyone stopped working on stuff that was due on Thursday. Nor have end-of-the-world cults come into being, focussed on September 10th. Although, isn’t that curiously close to September 11th? Oooh….

Human beings are hopeless at some things, and thinking rationally about disaster scenarios is right up there. Spotting when a coincidence is just a coincidence (and realising all the times the coincidence didn’t occur) is another problem for us. And NOT seeing patterns where there are none is a particular problem. These oh-so-human traits are probably behind 99 per cent of conspiracy theorising. I did a Canadian radio interview for Maritime Morning about 13 Things this afternoon, during which the host, Andrew Krystal ventured that most UFO sightings seemed to be in the same vein as medieval claims of miracles: to the people who were “there” they’re undeniable, but to those who weren’t, they’re astonishingly unconvincing. And there’s no way to make the two camps move any closer together.

In fact, it’s been scientifically proven that, given two extreme camps, a new piece of independent evidence divides them further. In the 1930s, two mathematicians, Frank Ramsey and Bruno de Finetti, independently derived results that showed scientific reasoning is mostly an exercise in reinforcing your prejudices.

You could interpret that to mean that we shouldn’t trust physicists’ assurances that the world isn’t about to end. But I’ve met Michelangelo Mangano, the man who worked on CERN’s security report on the black hole scenario and, honestly, there’s nothing suspicious about him at all. His time has, essentially, been wasted by a few people determined to deny all reason.

Somewhere common sense has to prevail. There’s nothing going on in the collisions at the LHC that isn’t going on just above our heads as cosmic rays (particles from the sun) pummel Earth's atmosphere, and there's plenty more violent stuff going on elsewhere in the galaxy. Despite millennia of this, no one and nothing has been eaten by collision-created black holes yet (or is that something else they’re not telling us?). New Scientist’s Valerie Jamieson put it well in a blog on the NS site (I got this via Cosmic Variance, another bastion of reason):

“Scale the cosmic ray sums up to cover the 100 billion stars in the Milky Way and the 100 billion galaxies in the visible Universe and you find that nature has already made the equivalent of 10 to the power of 31 LHCs. Or if you like, 10 trillion LHCs are running every second. And we’re still here.”

Valerie’s in Geneva at the moment, so if she’s wrong, she’ll be the first to go, of course...