The Singularity is Far Computing NatureArtificial Life, Virtual Worlds & Simulation Bruce Damer for the Singularity University August 3, 2010
• I. The Birth of Computing (and the Von Neumann Bottleneck) II. The Birth of Visual Computing and Virtual Worlds (still running through the Von Neumann Bottleneck) III. State of the Art in Simulating Nature (Physics) for Space and Chemistry IV. Computing Nature (?) Discussion
III. Computing Nature (?) Can Von Neumann Do It?
Conventional vs Natural ComputationSystemic Computer model by Peter J. Bentley, UCL, Digital Biology Group Conventional Natural Deterministic Stochastic Synchronous Asynchronous Serial Parallel Heterostatic Homoestatic Batch Continuous Brittle Robust Fault intolerent Fault tolerant Human-reliant Autonomous Limited Open-ended Centralised Distributed Precise Approximate Isolated Embodied Linear causality Circular causality Table 1 Features of conventional vs Natural computation
Non-living natural world supports a massive number of parallel interactions but they are finite, bounded
Living natural world supports infinitely repeatable computations in a massively parallel fashion
E-coli, a massively parallel computing universe David S. Goodsell from The Machinery of Life
E-coli, a massively parallel computing universe
The complexity of Cytoplasm A cube 100nm on the side contains roughly: - 450 proteins - 30 ribosomes - 340 tRNA molecules - several mRNA molecules - 30,000 small organic molecules (amino acids, nucleotides, sugars, ATP etc) - 50,000 ions - remaining 70% is water - all in continuous interaction
Nerve cells: two orders of magnitude more complex than e-coli
So can any kind of (Von Neumann) machine simulate a whole cell? Definitely not Low level approximations (overhead) How about a lot of these? Perhaps… for the equivalentof a small volume of aqueous chemicals, Anton: 1 microsecond per month
You need this…. to originate and evolve complex life (and civilization)Penny Boston, CONTACT Conference 2009, NASA Ames
Question: What is the computingarchitecture and cost of simulating a single neuron at the molecular dynamics level? Answer: This is beyond the current and probably subsequent two or threegenerations of supercomputers, even those dedicated to MD simulation.
Result: Even excluding the “non informational/maintenance” parts of the simulation of a neuron, the high fidelity modeling of a single neuron is still a substantial computing challenge.Therefore concepts of a Singularity as derived fromscience fiction (Vinge) remain wholly in the realm of science fiction.
So how to map this computer onto this one? Perhaps……toil for a number of decades toward a most minimal type of “Singularity”, an Artificial Origin of Life
The EvoGridAn “artificial origin of life” in cyberspace in this Century
Origins of Life: Archaean to Cambrian1997: Digital Burgess - quest for life’s algorithmic origins in the “Cambrian Explosion”, Biota.org
“Soft” Artificial Life Through the Ages:field named in the 1980s, progress through the 1990s, 2000s Early exemplar: Karl Sims’ Evolving Virtual Creatures (1991-4) Evolving Virtual Creatures by Karl Sims Inspired a generation of Soft Alife developers in the 1990s-2000s
Resources and Acknowledgements & DiscussionProject EvoGrid at: http://www.evogrid.orgProject Biota & Podcast at: http://www.biota.orgDigitalSpace 3D simulations and all (open) source code at:http://www.digitalspace.comWe would also like to thank NASA and many others for funding support forthis work. Other acknowledgements include: Dr. Richard Gordon at theUniversity of Manitoba, Tom Barbalet, DM3D Studios, Peter Newman, RyanNorkus, SMARTLab, Peter Bentley, University College London, FLiNT,Exploring Life’s Origins Project, Scientific American Frontiers, DigiBarnComputer Museum, The Shelby White and Leon Levy Archives Center,Institute for Advanced Study, Princeton, NJ, USA, and S. Gross.