Hive Mindstein

David Basanta’s blog has an interesting thread (quite a few of them actually).  Here’s the setup but you should read the original post, including the Wired article:

Apparently, some people are seeing some potential in cloud computing not just as an aid to science but as a completely new approach to do it. An article in Wired magazine argues precisely that. With the provocative title of The end of theory, the article concludes that, with plenty of data and clever algorithms (like those developed by Google), it is possible to obtain patterns that could be used to predict outcomes…and all that without the need of scientific models.

If I told you I had a massively parallel device which takes in huge amounts of raw data and finds patterns via computation to make predictions, would you be able to tell if I was speaking of a computer cloud or a human brain?

Anderson’s postulate about the end of theory is as specious as the claim that all good science is reductionism only, but for opposite reasons.  Statistical data-driven discovery of the kind Google is doing is in fact exactly a process of model-building (aka theory), no more, no less.  And when the human brain “creates” theory, it is doing so in the context of empirical data, only a small fraction of which is acknowledged explicitly.

Understanding cannot be gleaned from brute-force data-mining alone.  There are inherent complexity barriers that even an atoms-in-the-universe sized computational cloud could not crack.  On the other hand, theory without strong correlation to the empirical is just math.  In a tale of two sciences, theoretical physics has been negligent on the latter count while biology on the former.  Both pendulums are swinging back toward the center now out of absolute necessity.  String theory isn’t going to be a grand unification any more than DNA sequencing is going to cure cancer.

The interesting question that cloud computing raises is the limits of the (individual) human brain as a tool for scientific advancement.  One could argue that the reason we are stuck on so many big problems is because we are running up against the limitations of the human brain from both a bottom-up computation perspective as well as a top-down analytical reasoning perspective.  If this is so, then the importance of technology (like cloud computing) goes up and the likelihood of there ever being another Einstein goes down.

We should not forget though that human minds can be organized in coordinated activity to yield better results than any one individual.  But there’s no such thing as truly automated discovery either.  Not because of some special property of the human mind but rather because “truth” is only relevant within the context of a value system.  Whether a cell is cancerous and needs to be killed depends whether you are the cell or the multicellular organism.  And even if you believe in absolute truth, a human value system matters from a practical standpoint to sort out relevant truths from the infinite possibilities.  Ultimately, advancement in science will rely on computation beyond the ken of any one human, but it will be directed by humans who care about the discoveries and have a point of view on why they matter.

  • Hi Rafe,

    Nice post and good point, there might be limits to human understanding. There is probably a threshold of complexity beyond which a single mind will be unable to understand a physical/biological phenomena but I hope we will not yield to the temptation to produce ununderstandable models when a better one is possible. Finding correlations with models that we cannot understand can, undoubtedly, be useful, but it is in the best case, only a lesser kind of science.

  • gregorylent

    giving direction, and discovery, are the same thing … and context, pov, comes later