Saturday, March 27, 2010

When Scientists Go Bad

All that glisters is not gold. ~ William Shakespeare, The Merchant of Venice

Not long ago, there was a program on the Science Channel that should have been very good. It wasn't, because it missed the point, big time.

The program was about Hendrick
Schön, an apparently brilliant young physicist, who was earning the admiration and envy of physicists everywhere for research he was doing on organic transistors. Unfortunately, his papers were based on fraudulent data.

The show I saw was the second try by the Science Channel to report this story. The first time, they spent more time talking alternately about the coming economic crash that would be caused by the "failure" of Moore's Law and by the devastation that would be caused by the solution to the "failure" of Moore's Law, nanobots. In the second version of the show, it appeared that the only concern was our destruction at the prosthetic hand of the nanobots, because they didn't mention Moore at the outset. Unfortunately, by the halfway point of the show, they felt they had to resurrect the ghost of Moore (poetical allusion; he's still alive).

In short over half of each program was dedicated not to the subject of
Schön's fraud but to the mythical end of the world as we know it.

In case you're curious, the economy was going to crash because Moore's Law was about to fail, because technology had reached a tipping point where manufacturing methods would not allow the increase of the number of transistors on a chip. This bit of nonsense ignored the advent of multicore chips, allowing even a modest PC to have the power of multiple computers. It also ignored the research going on in quantum computing, where atomic states substitute for the switching of transistors.

In short, computing power is nowhere near its limit.

The supposed answer to the Moore's Law dilemma was organic transistors or, more descriptively, molecular transistors, which is what
Schön's research was about. The show took the idea of molecular transistors, using carbon-based organic molecules and went all Terminator-II-meets-the Matrix on us. A silvery-gray goo would end up devouring all humanity, so either way, we're all gonna DIE!

This is what passes for science programming these days.

What the geniuses who put these shows together failed to recognize was that there were several huge issues staring them in their collective faces, none of which involved Gordon Moore or gray goo.

For example, how did
Schön manage to get so many papers accepted by Nature and other prestigious publications without the peer-review process catching his data shenanigans?

Or, why did Bell Labs, where
Schön worked, not have some sort of oversight that might have prevented Schön from creating this mess in the first place? And what of his co-authors? Each paper had co-authors, yet none of them had a clue that they were aiding and abetting a fraud?

Or, what do
Schön's actions say about the pressure to publish and the apparent unwillingness of other scientists to actually verify the work? And how many other researchers are doing this stuff?

Or, and this is the ultimate question, why did
Schön do this in the first place? Oh, they give a little lip service to the subject, but it doesn't appear that they talked to anyone who really could offer an opinion.

Oh, and what ever became of
Schön? There is not a word on this subject (he's working for a German engineering consulting firm, in case you're curious).

Fortunately, someone has addressed these questions in a book that I've just started reading, Plastic Fantastic: How the Biggest Fraud in Physics Shook the Scientific World , by Eugenie Samuel Reich. Ms. Reich, a science reporter has dug into
Schön's past and his thinking. In just the early sections I've read, it's clear that Schön really didn't think he was doing anything so wrong because he believed that the underlying science was correct. The inconvenience of the experiments not bearing the theory out was dealt with by cooking the books.

But the book also deals with the larger issues of scientific integrity, the pressure to publish to get grants, the culture in research that would allow something like this to happen, and so on. All of this is fascinating stuff that the Science Channel could have spent a little time discussing if they had bothered to talk to any of the people Ms. Reich has interviewed (like
Schön's boss at Bell Labs).

But that wouldn't be as much fun as doing animations of gray goo oozing over the landscape.









Sunday, March 14, 2010

Happy Pi Day!

Some people know pi to trillions of decimal places.. actually they don't, but that's beside the point.~ Lee Houghton

Hopefully, you have discovered this little missive on March 14, for that day is known to many as Pi day. For those of you who are totally dense, the mathematical constant pi, which is the ratio of the diameter of a circle to its circumference, is usually given as 3.14. The more astute among you might recall it as 3.14159 (hence the time of this post).

Actually, pi is an irrational number. This doesn't mean that pi is argumentative or illogical. It simply means that it can't be expressed evenly as the ratio of two integers. In other words, calculate pi however you want to, you'll never get to a point where you're done resolving it. Pi has been calculated to gazillions of digits, mostly by people who have a lot of computer time and no girl friends.

Just to give you an idea, here's a relatively short value for pi: 3.14159 26535 89793 23846 26433 83279 50288. The spaces are there for readability. If you want to impress your friends, you can rattle this one off at the next party you attend. Of course, it'll be the last party you attend, but you'll have had your moment in the sun.

The funny bit is that, after 3.141592, you could probably rattle off any group of digits without fear of contradiction. And you still wouldn't get invited to any parties.

Assuming you've got anyone left listening to you, you could wow them with something like this:

At the twenty-seventh decimal position, you'll find the number 27, flanked by 3 and 9. This is explained by the person quoted above, Lee Houghton, in a comment at Pi Land. Even Mr. Houghton admits he needs a hobby. As do people who regularly frequent a site called Pi Land.

If you want to disseminate some really arcane facts, read this article. I guarantee people will avoid you for weeks.

People love to get all goofy about calendar dates that seem significant, like June 6, 2006. This, of course, can be written 6/6/06, which got some of the nuttier folks around going off about all that number of the beast (666) nonsense. Well, besides 666 being silly, relating 6/6/06 to the number makes no sense because it's got a ZERO in it. I know of no Biblical significance to 6606. Of course, there was a 6/6/6 in the year 6, except that no one was calling it the year 6 back then (that would come much later when some monk or another calculated when Jesus was born --incorrectly, as it turns out). Besides, Revelations hadn't been written yet, so had people been calling it the year 6, they still wouldn't have related any significance to it.

Personally, I think we should have more number holidays, but most people don't bother to remember the math they learned, so making a fuss about February 7, xx18 being related to e, the base of the natural logarithms probably isn't going to draw much of a crowd.

The square root of 2, which relates the value of the diagonal of a square to its sides, is a friendly 1.414 (or 1.41421 35623 73095 04880 16887 24209 69807 to the more precision-minded). But, like e, it's a once-a-century holiday.

Let's face it. The real reason we remember pi is because of that old, old joke:

Father: What did you learn in school today, son?
Son: We learned pi r2.
Father: Well, that's stupid. Pie aren't square. Pie are round.

Sad, but there it is.

Friday, March 12, 2010

Nothing Succeeds Like Failure Redux

The significant problems we face cannot be solved at the same level of thinking we were at when we created them. ~ Albert Einstein

I first wrote about SpaceX and its Falcon about 3 1/2 years ago, in a little pastiche that got me into a big go-round with some Space Frontier Foundation folks. It was a passing reference to the Ansari X prize and the fact Falcon 1 had gone blooey in its first launch.

Six months later, SpaceX was trumpeting that their latest failure was in fact a success. Political spin doctors have less nerve than Elon Musk. When the rocket failed because the second stage stopped firing. I noted that this event involved a definition of success with which I am not familiar. Oh, and the cost of this to U.S. taxpayers was a tidy $278 million. I have nothing against using tax dollars to explore the cosmos; in fact, I favor the idea. I'm just against throwing money away.

Apparently fearing others might have the same idea, SpaceX decided that it wasn't sufficient to merely classify a failure as a success. They now declared a rocket that had never completed a flight "operational." As I said, at the time, this is the hubris medal with an oak leaf cluster.

A year later, we find, thanks to the incredible track record of SpaceX (the rocket hadn't killed anyone yet) the government had awarded "an indefinite contract" to the company for their services to deliver supplies and people to the International Space Station.

As if the hilarity hadn't gone far enough, SpaceX tried again a few months later(I mean you gotta get at least one of these things to fly). This time the second stage once again stopped firing. It stopped firing because the first stage whacked into it, due, apparently, to a lack of understanding of how stage separation should be timed. Of course, Elon Musk once again termed this a success.

Remember, SpaceX is trying to build a rocket to carry people. One wonders how their next-of-kin would feel about this sort of "success."

Well, somewhere along the line, SpaceX finally got a Falcon 1 to work. Based on their one real success, they felt ready to move on. Falcon 9 (which is essentially 9 Falcon 1 engines strapped together) is sitting on a pad at Cape Canaveral doing its best to live down to the reputation of its predecessor. In this "success", the engine aborted "nominally", meaning it quit working, but it did it nicely.

At least it didn't blow up. Presumably, that comes later.

The President has already decided that using Robert Goddard technology to go to the Moon doesn't make sense. Unfortunately, he and his advisors haven't come up with anything better yet. Congress is already talking about extending the shuttle. The only good news is that the idea of scrapping the ISS in 2015 (President Bush's bright idea) seems to be losing steam. I don't like the ISS, but putting all that time and money into a project only to bring it down after less than five years of full operational status is a waste of monumental proportions.

I don't know where the space program is going, which is okay, because I don't make the decisions or spend the money. Trouble is, the President, Congress, and, worst of all, NASA don't seem to know either.

Let's hope they at least remember to get everyone off the ISS before they shut it down.