See http://www.kurzweilai.net/why-evolution-may-be-intelligent-based-on-deep-learning

My comment there:  "I recall coming to essentially the same conclusion in the late '60's after reading Koestler's "The Act of Creation," or perhaps it was "Janus."  I think it was referred to me by the Nathaniel Brandon Institute. As I recall, Koestler did not believe in junk DNA, for good reasons having to do with basic information theory. 

I agreed, but, in part, my agreement rested upon a an unanswered puzzle - how, in the face of random mutations, could millions of codes be stable enough to sustain a cellular structure over many generations?  The most conservative analysis that I could muster was orders of magnitude away from reality.

I concluded - as did Koester, I believe - that there had to be level of meta-programming, an internal OS that controlled and directed the DNA/RNA machinery and included some kind of sophisticated error checking.  (Had I but known of Mandelbrot's work..) Koestler also treated the genetic expressions as results of an internal program similar to a computer OS.  Once again, however, orthodoxy and the ruling paradigm ignored Koestler's work.  Not being a laboratory scientist, how could he know anything?   Koestler's model allowed for group evolution!  What heresy!!

So, important as this research appears to be, it is still overshadowed by  Kuhn.  We still have not solved the overarching problem of paradigm politics.  Just as the religious fervor against the notion of group selection dominated anti-aging research for about half a century, resulting in massive misallocations of R&D and the failure to recognize key elements in the aging process, so the recent research discussed here in: http://www.kurzweilai.net/no-longer-junk-dna-shedding-light-on-the-dark-matter-of-the-genome should probably have been done long ago, were it not for the ruling paradigm problem.

Just as evolution has itself evolved to promote its ongoing success - as illustrated in the male/female dichotomy, in which males become the primary selection mechanism (expendable and with only one copy of critical genes) - so we really ought to focus on how to use the scientific method on itself. 

This is a public goods problem, of course, bringing to mind Hardin's "The Tragedy of the Commons."  Where would the funding come from for a major effort to identify flaws in the scientific process, analyze them to find the key mechanisms, propose solutions and clinically test them? And then, how would such a research team sell their results to the establishment research community?

Could we make a game that concretised abstract systems?  One that rewarded more sophisticated and fundamental systememes as such?