Wednesday, February 6, 2013

Is Scientific Genius a Thing of the Past?

"Is scientific genius a thing of the past," asks Kate Shaw in Ars Technica, and points our attention to a nature paper by psychologist Dean Keith Simonton:
Today, according to Simonton, there just isn’t room to create new disciplines or overthrow the old ones. “It is difficult to imagine that scientists have overlooked some phenomenon worthy of its own discipline,” he writes. Furthermore, most scientific fields aren’t in the type of crisis that would enable paradigm shifts, according to Thomas Kuhn’s classic view of scientific revolutions. Simonton argues that instead of finding big new ideas, scientists currently work on the details in increasingly specialized and precise ways.

Where have all the geniuses gone?

Obviously, we do not have a cult of geniality in the sciences (or rather, in the ranks of the spectators of science) any more, while scientists, by and large, have not gotten stupid. So where do the big thinkers and paradigm shifters hide? Wait, no, obviously they don't. Garret Lisi might spend most of his days surfing, Grigori Perelman prefers to stay with his mom, Steven Kosslyn tends to Bay Area startups, Noam Chomsky is busy being hated for his liberal views, Ben Goertzel dithers in Hong Kong, Cristiano Castelfranchi might be caught in the debris of what was formerly known as Italy's academic system, Jerry Feldman and Marvin Minsky have effectively retired. On the other hand, many others struggle to escape postdochood, for in times of reduced public spending, tenured positions for out-of-the-box thinkers are scarce in many disciplines, or they have chosen obscure corners of academia that will forever deny them fame. Or they choose a career in business instead and work at Google and Amazon.

But the Einstein level geniuses are out there, and many are highly active and efficient. Some, like George Lakoff, Stephen Wolfram, Vidala Rachamandran are even moderately famous. Perhaps the problem is not the lack of genius, but its overabundance? In a highly connected world like ours, the many voices of genius might just be drowning each other out!

What really irks me in Simonton's statement is the bigger issue: that most current sciences are not in need of new paradigms. How he can entertain that notion really is beyond me!

Many sciences are in dire need of a revolution

To give just a few examples of severely broken sciences:

Simonton's own discipline, psychology, is supposed the science of the mind. However, psychology is almost completely a-theoretic. Yes, in case you did not know: psychology has no theory on how the mind works, nor does it endorse working on one. Psychology has somewhat recovered from the dark age of behaviorism, but so far it is still caught up in a very narrow experimental paradigm. Connecting the dots and building theories with free variables, as it is common in physics, is unthinkable to someone who has received the crippling conditioning of modern psychological methodology. Psychology needs bold overarching theories, which will probably have to rest on computational models.

Systemic sociology has been killed off by the influences of continental philosophy and Maturana (who inadvertently poisoned Luhmann's mind and thus turned the idea of a systemic modeling of society into stillborn wreckage). Many computer game companies have better abstractions of societies and social interaction than academia. A new systemic sociology will have to start almost from scratch.

Economy is an abomination, almost entirely perverted by political interest groups, and quite thoroughly detached from empirical realities. To the casual observer, the failure of economists to predict and come to terms with the current economic catastrophes is almost as flabberghasting as the fact that these guys are allowed to proceed as if nothing was wrong. Separating wheat from chaff will become a very difficult endeavor, since actual economic interests are at stake.

Philosophy of mind? This is what remains after everybody with actual methodological knowledge and contemporary scientific training has left the room. In its current form, it is about as useful as a philosophy of digestion would be, especially if it were done by people without current knowledge in chemistry, biology and cooking.

Physics has made tremendous progress until about 50 years ago, and has since shown somewhat diminishing returns in its attempts to build a unified theory of everything. Perhaps Wolfram's idea of enumerating and testing the possible world formulas would be better than the current attempts to stitch incompatible formalisms together?

The science and practice of programming is still in its infancy and reflects our poor understanding of how the mind works. Programming languages and techniques do not differ with respect to what they can make the computer do, but in how human programmers can bend their minds around the problems that they try to convey to the computer. One of the first slightly decent programming languages, Lisp, arguably still represents the state of the art.

Artificial Intelligence has reaped many of the low-hanging fruits in the sixties and seventies, and lost its direction. Today it builds applications, appliances or (quite successfully) tinkers in applied mathematics. But to get back on track to build an artificial mind, it will need to gather around a genuinely AI related productive paradigm again.

Cognitive Science, the attempt to integrate psychology, neuroscience, AI, linguistics, philosophy and social sciences to take over where AI has left off, has quickly turned into a bunch of incompatible methodologies competing for the same funding bucket. That fight has been mostly won by its most expensive parent, neuroscience, and now cognitive science spends most of its budget on putting people in brain scanners, which seems to be about as helpful as the attempt of understanding flight by looking at feathers through microscopes.

Does the above sound pessimistic? I think not! I think it means that science is full of impending revolts and revolutions. Science needs genius, and even though genius does not need science as much as it did in the past (I hear that Google is still hiring), big thinkers are bound to be gathering at the foundations of many established fields and topple their rotten structures, as they have done in the past.