Soft Computing Lab. Dept. Computer Science Yonsei Univ. Korea Self-replication from random parts 30 Mar. 2010, Keunhyun Oh Saul Griffith, Dan Goldwater, and Joseph M. Jacobson MIT Media Lab 0 Nature, vol. 437, pp. 636, 2005.

Contents Overview Related works –Key feature of biological replication –Previous studies Proposed methods Experiments Summary and future works 1/15

The autonomous self-replication Autonomously self-replication machines –not yet to acquire the sophistication of biological systems, which assemble structures from disordered building blocks The autonomous replication of complex systems from random inputs –A replication of a reconfigurable string of parts –Randomly positioned input components Components –Suitable Miniaturized and mass-produced –Constituting self-fabricating systems whose assembly is brought about by the parts themselves 2/15

3/15 Key feature of biological replication Selecting the appropriate building blocks(nucleotides) from parts (ex. DNA) –Parts randomly and continuously distributed in its environment Correcting errors made during copying The efficiency –Enabling biological systems to generate exponential numbers of accurate copies of themselves as a function of time To create these properties –Autonomous acquisition of randomly distributed building blocks –Carrying out error correction during the copying process

4/15 Previous studies A scheme for the autonomous self-replication –A simple 2-bit mechanical string outlined almost half a century ago Using structured inputs has since been achieved Including a self-reproducing machine that relies on a well- ordered supply of its building blocks

5/15 The complexity of a given structure The bit length describing the configuration of parts – in this case, a 5-bit string Ɛ –the error per addition(arising from random input) of each new building block in the copied string (1- Ɛ ) n –The yield for replicating an n-bit string –Exponentially small for complex (large n) systems ( Ɛ = 0.5, n=5; the yield is about 3% in the case described here)

6/15 Error correction For complex structures to be copied accurately from random inputs A process in which a linear increase in resource leads to an exponential decrease in error rate DNA replication –the polymerase enzymes responsible for copying may also check each recruited nucleotide base for correct complementary base-pairing with the DNA template strand –if the incoming base does not fit, it is removed by the enzyme’s exonuclease domain.

7/15 To implement error-correcting replication A set of programmable electromechanical components –Run as a 7-state, finite-state machine –the components can be reversibly latched and unlatched in response to nearest neighbor communications –These parts interact by floating on a two-dimensional air table on which motion is random Self-replication of this sequence –A result of a random part latching on to the seed string –the part is queried for self-similarity and proper position in the growing replicant –subsequently it is either permanently latched or released according to an embedded rule

8/15 Growing Machines S. Griffith, 2004

9/15 6-state machine S. Griffith, 2004

10/15 Electromechanical units

11/15 An example of latching each other

12/15 Self-replication of a 5-bit string Figure 1 shows a series of frame shots that start with a single- seed string (coloured in a green, green, yellow, yellow, green sequence).

13/15 Result The kinetics of these processes are exponential until they become limited by the supply of parts(Fig. 2).

14/15 Another formation

15/15 Summary and future works Summary –The autonomous replication of complex systems the ability of the DNA template to select the right building blocks A set of randomly scattered parts The ability to correct copying errors –developing machines for the autonomous self-replication of a reconfigurable string of parts from randomly positioned components. Future works –Machines will be more miniaturized –Possible to create a general system self-replicating and programmed to self-fabricate into complex structures that run with exponential kinetics.

Thank you 16