technology increases exponentially – in fact, the rate of acceleration increases exponentially
can you predict the future?
certain features of information technology are highly predictable
crazy graph of countdown to singularity
similar to
events are getting quicker exponentially
3 biological:
change of skull for weaker jawbone but bigger brain
more of the brain for cereberal cortex
thumb joined moved one inch
then social, now technological
took many lists of key event, and they all fit this graph
Moore’s Law – the fifth paradigm finishing in 202x, creates pressure for research for the sixth paradigm (3d nanocomputing)
hit the same power of supercomputers and the human brain in 2013 (all human brains bt 2040)
the biotechnology revolution – the intersection of biology with information technology
new paradigm of rational drug design
communications doubling (everything) every 12 months
if you look on a linear graph, it appears the Internet came out of nowhere, but is apparent on a log graph
nanotechnology – blood-cell sized machines – happening now
biology is pretty suboptimal, everything built of proteins
nano versions of red blood cells are hundreds of times more capable
neural implants for parkinson’s disease – FDA approved, download new software from a computer outside the body
exponential gains in understanding how the brain works – reverse engineering the human brain
simulation of the cerebellum
how complex is the brain? well, it’s contained in the genome – 50 million bytes? manageable
scenarios –
2010: computers disappear
embedded everywhere
full immersion virtual reality
augmented real reality
2029: 1000 dollars of comuting power is 1000 times the human brain
computers pass turing test
nanobots provide neural implants, full immersion virtual reality from the nervous system
expansion of human intelligence
http://www.kurzweilai.net/pps/tretc
comments
I think I would have been quite annoyed with this talk. The comments about biology being sub-optimal are silly. There is an implicit assumption about what “optimal” is. Can we make nanotech red blood cells using the same amount of energy as an animal? Can that nanotech red blood cell defend itself against pathogens, adapt to numerous special conditions, be recycled to create new cells? Can it evolve over generations, with the help of the body containing it, to become better? And what kind of red blood cell is he referring to – human, whale, albatross? And in what way are the nanotech ones better? Clearly the information contained in the genome is of a manageable size, but is the system that genome describes manageable? No it isn’t. It isn’t even (yet) comprehensible. I think his optimism is based upon an incomplete understanding of the things he’s trying the replicate artificially. As for the precise point in doing this anyway – that defeats me.
— Ben 3.10.05 #
I’ve just read through the presentation. I think the most revealing element is the analogy of evolved design of organisms with human design of electronics (circuits, block diagrams). Although it is compelling to imagine this is valid, it is also wishful thinking with little to back it up. These design approaches work well for us as we have a clear final goal, and once the machine is built it doesn’t have to change itself in a battle to survive. For organisms, the “goal” is survival, and there really is no knowing what you need to change or do in order to survive. The changes are undirected, and any bad changes are eliminated. This kind of pressure favours certain designs, which include redundancy, fault-tolerance and error correction at every level. It is the kind of design which human designers flinch from, as it appears extremely wasteful (and it
is, when you’re not having to deal hundreds of unexpected occurrences). Efforts to fit biological systems to neat “electronic circuit” or “computer program” type designs have, for the most part, failed (in my limited experience). The horrible thing about biology is that there are exceptions to every rule, and sometimes the exceptions are in the majority.
— Ben 3.10.05 #
<< previously
home
furthermore >>
