Wednesday, September 06, 2006

Depressed by Innovation

Reading Terence Kealey's The Economic Laws of Scientific Research, I stumbled onto an interesting passage concerning the popular (Keynesian) view on what caused the Great Depression.
But Roosevelt and the Democrats believed that improvements in productivity had caused the depression by throwing people out of work....If Hoover supported science, they argued, then science had to be a bad thing....Much of the nation shared that believe and, if there was an understanding that science might improve living standards in the long run, there was also a widespread believe that the time had come for a moratorium on research. [Emphasis added]

Apparently Science magazine in 1935 (vol. 81, p. 46) published some popular thoughts concerning technology. Kealey presents this enlightening example: "the physicist and the chemist seem to be travelling so fast as not to heed or care where or how or why they are going. Nor do they heed or care what misapplications are made of their discoveries."

Science is a bad thing? Technology moving too fast? A moratorium? Let's look back to the mid and late 30s and see some of the things these elitist were working on. 1937 saw the first blood bank. 1936 gave us the first tv broadcast with modern definition as well as radar's first working model. Enrico Fermi--and others--first studied nuclear fission leading, in part, to its discovery in 1939.

Don't let anyone tell you science is moving too fast or it should be slowed or scientists should be held back. It's easy to look to the past and defend what we might have missed. It's much harder to protect an unknown today so it may mature into tomorrow's cherished revolution.


Jenny said...

Yes…Science and technology is moving too fast. Look at all the jobs lost by blood-letters. Someone should definitely do something about that.

Anonymous said...

I will never say science is moving too fast, but I will say technology should be a bit more careful. Some decisions were made with internet security in the early days that made spam and spoofing a big problem today. Unfortunately by the time we realized what happened, it was too late. Think of it like if we learn cars are safer if the steering wheel is in the middle (like when you play a racing game). We're so used to the wheel being on one side, it won't change.

How often have we had to put up with patches, recalls, defects and all the other fixes to problems brought on by the rush to get the latest and greatest out the door?

Then we have the legal issues like cyberstalking and security issues like terrorists hiding messages in image files and moral issues like cloning. The problems are a lot more serious than before when Edison bragged about making electricity so cheap only rich people would burn candles.

We can do more with science and technology than ever before. The rewards are great, but so are the risks. What's wrong with asking for some caution? Especially since taking some time to work out the kinks means less time and resources spent fixing them later.

David said...

I think the more interesting question is how often do not have to worry about recalls or defects? Even computer patches are pretty rare considering how many things could go wrong. A lot of people see some problems and thinks that's too many but never consider how often things work fine, an understandable error but an error nonetheless.

Now it's true that everyone wants stuff to work all the time. And so companies test their product. And test it. And test it. It's naive to point to problems and proclaim they clearly don't test it enough. While costs can be saved if all the bugs are worked out, detecting those problems beforehand increase rapidly. A fundamental problem is easy to detect. A bug that only becomes a complication because it interacts with another product or only appears when employed at the large scale or has a low chance of showing up is incredibly costly to find. That's why most patches/bugs/recalls are based around those kind of problems.

The very fact that these corrections exist demonstrates scientists are taking responsibility for their actions. To claim otherwise because they didn't get it perfect the first time around doesn't treat them as human but rather lazy gods.

Anonymous said...

I think you're making the error of confusing science and technology. One goes with the other, but they are seperate. Science is research and learning new things. Technology turns that into things for people to use. Some scientists develop technology and some technologists do research in science. But one does not mean the other.

Now with computers, you're venturing into my territory. I develop computer programs. I'm not a scientist. We do updates and bug fixes every few weeks. Microsoft releases their own fixes at least twice a month.

Microsoft is mostly technology, it hasn't done much acutal science or inovation. They have an interesting damn if you do damn if you don't situation. They decided to make their OS backwards compatable. You can take Word 95 and in theory use it with Vistia. Nice if you don't like buying new software, but they decision not to abandon obsolete code leaves doors open for hackers. With their patches and service packs, Windows became more and more complex and harder to improve without causing more problems. It's become a hodpodge of DLLs and EXEs offering so many linked features one bad program can do horrendous damage.

But if they ever stopped and drew a proverbial line in the sand to prune the deadwood, who knows how many programs will stop running? And it's only going to get worse as time passes. It may be extremely costly to find new bugs, but unless Microsoft does something drastic, it won't have a choice.

One person can miss a flaw, the odds of two people missing the same flaw are smaller. The more people doing QA (quality analysis), the fewer bugs make it into final release. I can live with an OS as complicated as Windows needing an occaional fix, same is true of online games and so on. But what about an important system like airport control tower I don't think it's too much to ask that it be tested extensively and not have a patch be released a month later.