Technology driven science

Ever since around Galileo's time, which was also the origin of modern science, empiricism replaced deductive reasoning as the core of basic knowledge.  This changed worldview was stimulated by many things, but largely by instrumentation.  In particular optics drove fundamental realizations about the world that could not have been made previously.  One thinks of the vast new worlds of facts revealed by telescopes and microscopes, but other things came along as well, such as the discovery of vacuums, basic findings in chemistry, anatomy, and geology among others.  Navigation with the aid of improved astrolabes and, especially, reliable clocks, opened the world to faster, safer navigation.  Steam power led to trains and improved mining and factories.

Instrumentation led to a wealth of new data, and that stimulated thinking both in theoretical science (e.g., gravitation, the sun-centered planetary system, and so on).  Partly this was driven by the desire to control commerce and global political power.  Engineering and 'science' became more part of each other.  This led to understanding the law-like nature of Nature, and stimulated evolutionary and even social science thinking.

Today, we are driven even faster--much faster--by technology.  Technology people to dream of having an advantage over research competitors, and companies push their gear to feed their own interests but also those of their academic and industry customers.  Genetics exemplifies these trends.  But are we  paying a fair price for the gains?

I'm at a meeting in bioinformatics, in Poland right now, and the new DNA sequencing and other technologies feature prominently.  The technologies are mentioned almost as much as the results, and talk after talk is about what 'can now be done' with the latest computer or sequencing powers.

Clearly, as we've said countless times here on MT, we are not getting the results we dreamed of when it comes to genetic causation and the prevention of all known human ills.  We are just as clearly learning new things--such as the discovery of all sorts of RNA that wasn't supposed to be there according to previous genetic 'theory' (one gene, one protein, for example).  Were it not for the misleading hype by which the system is driven, one might feel less like complaining.  But even advocates realize that these tools have neither revealed a new theory of life nor reaped their promised miracles.

One problem, widely recognized, is that the new technologies are pushed on the marked before they are really properly ready and battle-tested.  The drive, perhaps especially in the US but spreading around the world, to have the newest machinery, is largely responsible.  Don't wait for the stuff to work really well--get it now and get a jump on your competitors!  You can publish results and shamelessly acknowledge then, or at least later, that you know or knew there were plenty of problems, but it was, after all, only exploratory or pilot or tentative data.

One thing this impatience does, besides cost a lot, is fund the companies for years while they make their gear actually work.  New types of DNA sequencers are widely acknowledged to produce many errors on the sequence they report.  Extensive and expensive efforts, paid for by grants to a great extent, are made to use the stuff and document the errors and attempt to correct for them.  But they are not removed, and genomic and other data bases are now loaded with sequence and other types of data that have too many errors, and they are often inscrutable.  How do we decide what data to use, and what data we have to ignore (or ask the public to pay for again, now with updated technology)?

We keep the tech companies afloat by buying products that are flashy but not yet ready for prime time.  It's one thing to acknowledge that anything highly sophisticated will improve over time, but it's another to go in too far, too fast which is what we're doing.  In part, I think (though I'm no business expert of any sort!) that the companies are often start-ups that don't have huge amounts of up-front investment: they have to sell sooner rather than be more patient, because they're not building on top of other mainstream things that they sell.

It's a problem in this field, and it has many consequences.  Not only the cost, but the loading up of massive data bases with suspect data, tentative but erroneous conclusions that lead to large numbers of follow up studies by investigators eager to jump on bandwagons, follow the latest trends, or get better information on some important problem such a serious disease, are examples.

Better and more restricted focus, less 'omics' and more theoretical understanding and basic science, and a slowed down pace would help.  But these seem impossible in our current heated-up system, unless the funds dry up.  If that happens, it will be bad in some ways because some of these new technologies really are important, but it may be good if it forces us to think more before we act.