Thursday, November 10, 2011

Stephen Wolfram: Computation & The Future of Mankind 5 @Singularity Summit 2011

We continue with the lecture given by Stephen Wolfram titled, Computation & The Future of Mankind delivered at the Singularity Summit 2011.
The lecture will be in italics in white.  Our annotations will be regular type in orange.

You look at all these systems in the computational universe, you see them doing all this amazing stuff that they're doing, but is their purpose?  Sometimes you can look at a system and say what that system does is more economically explained by saying it achieves a certain purpose, than just that it operates by a certain mechanism.  But ultimately our discussion of purposes is a very human thing.  It is very tied into the thread of human history that runs through our civilization.
I'm quite sure that in time, for all practical purposes, we'll be able to fix our biology and keep things running forever. Actually I have to say that an intermediate step surely be various forms of biostasis. I am often amused at what happens with things in science. I remember cloning, where I asked a zillion times, why mammalian cloning couldn't be done? There were always these very detailed arguments. Well, as we know, eventually this weird procedure got invented which made it possible. I very strongly suspect that the same kind of thing will happen with cryonics. I'm sure there is nothing fundamentally impossible about it. Right now, it's kind of a weird non-respectable thing to study, but one day some whacky procedure will be invented that just does it. And we'll immediately change all sorts of attitudes about psychology, about death and so on. But it's only just one step. In the end, one way or another, effective human immortality will undoubtedly be achieved and it will be the single largest discontinuity in human history.

When the thread of history is broken, it's very hard, even with something like Stonehenge, to know what its purpose was actually suppose to be.  When we look at the future, it's pretty obvious that more and more of our world can be automated at an informational level, and at a physical level.  One of my pet future projects that I don't think we're quite ready for yet is to turn robotics into a software problem.  I imagine some bizarre collection of little tiny identical objects moving around perhaps kind of rubicsbube style operating in a bit like cells in a cellular automaton.  A kind of universal modular object that configures itself in whatever way it needs.  I'm guessing that this can be done in a molecular level too.  I've actually got some pretty definite ideas as to how to do it.  But I don't think that the ambient technology exists yet.  We've still got a lot of practical infrastructure to be built.


When so much can be automated, even as I mentioned before, sort of creative things, it'll be interesting how economics change.  Things will still be scarce but so much, in a sense, will be infinitely cheap.  Then of course, there's us humans as biological entities.  One question one can ask is, how the kind of worldview that I have affects how I think about that.  People like me, have all this datum about their genomes for example.  But how does one sort of build the whole organism from that?  What are the kind of architectural principles?  Should we imagine that it's kind of like flow charts or simple equations, where one affects another, and we can sort of trace the diagram around.

I suspect that in many cases it's more complicated.  It's more like we see all over the computational universe.  Lots of cellular automata look so biological in their behavior.  We now know lots of detailed examples where we maps systems like cellular automata unto biological systems, whether at a macroscopic scale or microscopic one.  What we realize is that many aspects of biological systems are operating like simple programs, often with very complex behavior.  That means that all those phenomena like computational irreducibility come in.  Will some tumor like process in a biological system grow forever or not?  That might be like the halting problem for a Turing Machine.  [For those who are not acquainted with the halting problem, we include these two short videos.  If you cannot see the embedded video, here is the link:  http://bit.ly/vK4rVN.]

It might be formally undecidable and to know what will happen in a finite amount of time, we just have to simulate each step.  One can start thinking about all sorts of basic questions in medicine.  Here's an analogy.  Think of a human being, being like a big computer system.  System runs just fine for a while.  But gradually it builds up more and more crust, buffers get full whatever.  No doubt, the system has bugs too. Sometimes those get in the way.  Eventually the system gets to messed up, that it just crashes, it dies.  That's rather like what seems to happen with us humans.  Of course, with a computer system one can just reboot and start gain from the same underlying code, very much like the next generation humans can start again from more or less the same genome.

For humans, we have all this medical diagnosis, all these diagnostic codes for medical diseases and such.  We can imagine doing that for computer systems too, diseases of the display subsystem, diseases of memory management, trauma to the IO system, so on.  I think it would be instructive to really work this through actually.   What I know one will find, is that the idea of diseases is not really right.  There are all these different shades, things that can't really be described by parameters.  They have to be really described by algorithms.

When we run all the services of Wolfram Alpha for example, we have, by the standards of the computer industry, a very beautiful mathematical power dashboard showing us kind of an overall health of the system.  It all worked beautifully yesterday as the system turned on and so on.  But these things are very coarse.  I think it will be interesting how many more details we can represent well.

There's a lot of computational irreducibility lurking around. The very phenomena of bugs is a constant concept of computational irreducibility.  But the questions is, how should one best make a detailed dashboard that allows one to trace the path of problems in a big computer system.  It's sort of a model of what we should do for us humans, sampling all sorts of things with sensors or genomic assets or whatever.  Sampling what's happening then computing what consequences these things will have.

In case of big computer systems, we tend to just use redundancy and parallelism and tend to not worry too much about individual pieces dying. So we don't yet know too much about what we do and perform interventions. With humans we do care about each one, and we want to be able to do that.  It's going to be a tough battle with computational irreducibility.  No doubt we'll have algorithmic drugs and so on, where molecules can effectively compute what to do inside our bodies.  But figuring our the consequences of particular actions will be difficult, but sort of a lesson of computational irreducibility.

Still I'm quite sure that in time, for all practical purposes, we'll be able to fix our biology and keep things running forever.  Actually I have to say that an intermediate step surely be various forms of biostasis.  I am often amused at what happens with things in science.  I remember cloning, where I asked a zillion times, why mammalian cloning couldn't be done?  There were always these very detailed arguments.  Well, as we know, eventually this weird procedure got invented which made it possible.

I very strongly suspect that the same kind of thing will happen with cryonics.  I'm sure there is nothing fundamentally impossible about it. Right now, it's kind of a weird non-respectable thing to study, but one day some whacky procedure will be invented that just does it.  And we'll immediately change all sorts of attitudes about psychology, about death and so on.  But it's only just one step.  In the end, one way or another, effective human immortality will undoubtedly be achieved and it will be the single largest discontinuity in human history.

I wonder what's on the other side though.  So much of society and human motivation is tied up so much with mortality, when we have immortality and all sorts of technology, will we able to do almost anything?  The big question is what will be choose to do?  Where will our sort of overarching purpose come from?  It's sort of strange how our purposes have evolved from course of human history, how much of what we do today, in our intellectual world, or in virtual worlds, or whatever else, would seem utterly pointless and effortless to people from another age?  I think that computational irreducibility has something to say here.  I think it implies that there at least exists an endless frontiers of different purposes that can be built on one another.

In a sense computational irreducibility is at least one reason why history is meaningful.  If everything was reducible, one would just always be able to jump ahead and nothing would be achieved by all those steps of history.  But, even though new purposes and new history can be built, will we choose to do so?  Hard to know.  Perhaps for all practical purposes history will end.


I have a strange guess for at least part of this.  My guess is that when in a sense almost anything is possible, there'll be an almost religious interest not in the future, but in the past.  Perhaps it will be like the middle ages where it was the ancients who had the wisdom and instead in the future people will seek their purposes by looking at what purposes existed in a time when not everything was itself possible.  Perhaps it will even be our times now which will be of the greatest interest.  We are at that point in history where a lot is for the first time really getting recorded, but not everything is possible.  It's sort of a big responsibility for our time where we'll define the purposes for all of the future.  Perhaps it will be that way.

We will continue with our last installment and finish transcribing this lecture by Stephen Wolfram.

No comments: