EFTA02666879.pdf

From: Joscha Bach < Sent: Sunday, December 11, 2016 10:58 PM To: Jeffrey Epstein; Martin A. Nowak Subject: "Something big" vs "The end of science" Newton, Kant, Darwin etc. built new, essentially complete systems, =hrough which it became possible to understand a large part of the =orld. Perhaps the last one that was completely pervasive was =ositivism, in the early 20th century, together with a functionalist =ens. The century started out with the expectation of a systemic =evolution, which would especially manifest in physics. Since then, new =ystemic approaches have been found, especially computation and =ybernetics. But physics fizzled: Einstein's and Maxwell's universes =ould not be unified. Goedel ruined the party in mathematics. =ybernetics got lost in complexity theory. Minsky's Al and Chomsky's =inguistics seemed to have failed. And at the same time, the systemic =pproaches to society and governance failed. Kissinger killed the =hilean experiment of building a cybernetic economy before it even =tarted, the planned economies of East went bankrupt, and the market =conomies of the West stumbled from crisis to crisis. The systemic =deologies of communism, market liberalism and eurosocialism collapsed. Our fashionable postmodernist ersatz intellectuals, like Jarvis and =organ, tell us what most people want to hear: that systems are over. =nd nobody seems to say otherwise. Joi writes books in which he replaces =ystemic thinking with slogans: antifragility, whiplash, resilience, =indfulness, fluidity. Glorified versions of "muddling through". Personally, I disagree. The new systems are there, but our public =ntellectuals can no longer see them, because they lack the formal =raining to do so, and our scientists can no longer see them, because =hey tend to be too specialized to zoom out far enough. In my view, the =ain insight that drives the new kind of systemic thinking is the break =ith the mathematical tradition in the sciences. The book of nature is =ot written in mathematics, but in computation, i.e. an a much more =estricted script. When we apply this constraint, we get a new and more =roductive perspective on physics, causal structure, living organisms, =omplex systems. A big part of the new systems are accidental discoveries of Al. That =nformation is the basic building block of our universe (energy, space =nd matter are derived notions), learning and modeling are about =iscovering the eigenvectors and operators in feature manifolds, that =earning needs gradients pointing into the right direction, that =ausality is conditional state transition, that Bayesianism is the =eneral epistemological principle. Our public intellectuals are currently debating whether algorithms are =oo racist. They cannot muster enough attention to see that, due to the =ay in which minds model any domain, everything is an algorithm, =ncluding and especially evolution, i.e. the algorithm that has produced =s. If we accept that the universe and everything in it is fundamentally =omputational, i.e. can be characterized by the regularities in changes =n patterns of information, we get a unified frame of reference that =oes beyond the reach of every previous theory in history. This can =flow us to build an API for integrating all fields of knowledge and =ontrol. I suspect Barnaby can see much of that. Perhaps someone should slip him =ome coke to ruin his modesty, so he starts believing in his destiny for =reatness instead of quiet gardening :)=?xml version=.0" encoding=TF-8"?> <IDOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd"> <plist version=.0"> <dict> <key>conversation-id</key> <integer>60254</integer> <key>date-last-viewed</key> <integer>0</integer> <key>date-received</key> <integer>1481497228</integer> EFTA_R1_01918434 EFTA02666879  <key>flags</key> <Integer>8590195733</integer> <key>gmail-label-ids</key> <array> <integer>6</integer> <integer>2</integer> </array> <key>remote-id</key> <string>670254</string> </dict> </plist> 2 EFTA_R1_01918435 EFTA02666880