Natural Selection of Information – Can Google Help?

A friend of mine brought to my attention an article with the Top 5 Threats to science-based medicine. It mentions the usual state sponsored quacks, but the most interesting one is #5: New Media.

Of course, “user generated content” and unvetted health advice and claims are easily made by anyone anytime. And thanks to the magic of Google, a health claim need only be popular to be promoted.

So far, so familiar: it’s very hard to separate truth from nonsense when it comes to health information online.   Not that this is easy offline. It doesn’t help that even supposed authorities contradict each other and have conflicts of interest.

But what got me thinking is this bit:

Truth, accuracy, and scientific rigor aren’t always rewarded in this brave new world of digital influence. Being right has been uncoupled from being influential. The “wisdom of crowds” now decides what people see first when they attempt to educate themselves about health matters.

Because I just watched an episode of Beagle, my mind is primed at evolution theory, natural selection and in particular some of the painful consequences that can have to ordinary life. This leads me to the following question: in this evolutionary battle, can disinformation win from information?

This article and my intuition seems to suggest that it can, at least in the medium term (a couple of decades) . Not just for health: the problem exists in creationism, the housing bubble, politicians deciding on economic stimulus packages, the way we don’t properly deal with climate change, etc. There are always a few people who know the truth about these things, but these voices are not heard through the noise. As long as popularity and not truth is the strongest selection criterion for the stuff is in our collective minds, we are screwed.

I think the next challenge for Google and its competitors is to increase the relative strength of truth as a selection criterion. This is probably the most interesting aspect of Wolfram Alfa, but they rely on too much human labor. Could there be a lazy way (i.e. brute force, Google style) to spot the truth in a pile of nonsense, that does not rely only on popularity?

Take creationism, intelligent design and religion in general as an example. How does one become an atheist? The easy way – as in my case – is if your parents are (more or less) atheist, it’s generally accepted in your society and your teachers don’t try to convince you otherwise. But otherwise, you would have stumble upon The God Delusion or other sources with similar arguments. You would also need strong skills in separating weak or deceiving arguments from strong arguments. You need to able to find others who disagree with him and test their arguments. And you can not get around the social aspect either. The fact that he is a well known author does make a difference, if only because it means more people have taken an unsuccessful shot at disproving him.

Is this intelligence heavy process the only way to discover truth in cases where the majority of people are wrong?  Can it be replicated with current state of the art technology? And even if it could, would it stand a chance of becoming popular enough?

Until then, I fear that truth will be a luxury. Like the American Dream, it can be obtained by anyone with enough dedication, but most people won’t bother and many who do try will fail.

One Comment

  1. Pieter
    Posted December 27, 2009 at 23:19 | Permalink

    Natural selection could work if there was evolutionary pressure. Maybe with more easily testable hypotheses and emphasis on evidence based interventions (eg rapid learning systems in medicine), important information can penetrate through.
    What concerns me is talk of legislating for funding of alternative based therapies and giving non-pharmaceutically trained individiuals access to dangerous prescription based medication. http://j.mp/7FH600