Artificial emergence: visions and nightmares The visions of Drexler and others of molecular nanotechnology will be realised sometime. The vision is appealing — trillions of tiny autonomous machines creating useful products or cleaning up after us, using t

ofspace.Fromthisobservations,wedevisedathree-partarchitectureforemergentsystems[4].Wespeculatethatanycomplexsystemwithrequiredemergentpropertiescomprises,

elements(cells,nanites,ants,etc),describedbysimplestatebehaviouralproperties

detectableemergence(patterns,constructionsetc),describedintermsofthephysicalrealityorphysicalproperties

arepresentationorenvironment,supportingmappingsbetween,andpossiblyinteractingwiththeotherparts.

Wespeculatethatsomearchitecturalpartscanbedevelopedbydirectapplicationofexistingen-gineeringtechniques.PartofTUNA,forexample,isinvestigatingformalmodellingandre nementofaspectsoftheplateletsystem.

ConceptsandLayers

Havingestablishedanengineering-friendlyarchitecture,wearenowexploringmorethoroughlypatternswithineachpart.

CAmodelsemployonly“upwardcausation”:yersandcausationhavelongbeenstudiedinthecontextofemergence[1],and,whenwemovebeyondsimpleCAs,wetoo ndaneedtoincorporate“downwardcausation”.

InTUNA,wearelookingattheoreticaldesignsforarti cialbloodplatelets.OnesimulationusesastochasticCAwithupwardcausationtoarepresentation.Themodelisbasic;asplateletscluster,theymoveveryslowly,andriftsopenintheclustersastheymove.However,bymigratingthecontrolofclusteringandmovementoutoftheCA,weachieveahigher-layermodelthatiscapableofrespondingtoenvironmentalin uences;inkeepingwithreality,thehigherlayerexhibitsrelativelocationandrelativemotion.Atthelowerlevel,theCAupdate,stillsynchronisedattheCAlevel,occursinresponsetosignalsfromthehigherlayer.Anicefeatureofthismodel(underdevelopmentbyP.H.WelchandF.R.M.Barnes,UniversityofKent,usingoccam-π[2,6])isthatvisualrepresentationscanbeattachedeithertothelowlevel(toobserveabsolutelocationsandthewayplateletspassthroughthemovertime),orthehigherlevel(toobservetheemergentbehaviourofclustersofplatelets)[5].

TheFuture

Engineeringofemergentsystemsisclearlycrucial.Inourwork,wearecontinuingtoexplorelayers,causationandthethree-partarchitecture.Wearestartingtoconsiderargumentation,andwhatisneededtoarguethatadesignedorsimulatedemergentsystemwilloperatecorrectlyinanuncertainworld.Weplantoexplorethecaptureexistingknowledge(oremergence,complexity,argumentationetc)aspatterns.

References

[1]P.B.Andersen,C.Emmeche,N.O.Finnemann,andP.V.Christiansen,editors.DownwardCausa-

tion.Minds,BodiesandMatter. AarhusUniversityPress,2000.

[2]municatingMobileProcesses.InI.East,J.Martin,P.Welch,

D.Duce,andM.Green,editors,CommunicatingProcessArchitectures2004,volume62ofConcurrentSystemsEngineeringSeries,pages201–218.IOSPress,September2004.

[3]M.Hill,S.Stepney,andF.Wan.Thee ectofirregulargridsonthelongtermbehaviourofLife.

submittedtoECAL2005,April2005.

[4]F.PolackandS.Stepney.Emergentpropertiesdonotre ne.InREFINE2005,BCS-FACSRe nement

Workshop,ENTCS.Springer,April2005.

[5]FionaPolack,SusanStepney,HeatherTurner,PeterWelch,andFredBarnes.Anarchitecturefor

modellingemergenceinca-likesystems.submittedtoECAL2005,April2005.

[6]municatingmobileprocesses:introducingoccam-pi.InA.E.

Abdallah,C.B.Jones,andJ.W.Sanders,editors,25YearsofCSP,volume3525ofLNCS,pages175–210.Springer,2005.Toappear.