Creativite Artificielle Paradigm Shift 2011
Pour ParadigmShiftMeetings2011 session ~10h20-11h20
Pourquoi la creativite est-elle centrale dans la recherche scientifique comme artistique ? En quoi une intelligence artificielle beneficierait-elle a etre creative ? Comment pourrait-elle acquerir des mechanismes arrivant a un tel resultat et a quel prix ? Si de tels mechanismes etaient efficacement implementes dans une machine, quelles seraient les consequences sur le plan intellectuel, social, humain ? Cette presentation aura pour but de clarifier de telles questions et d'essayer d'esquisser des reponses d'apres l'etat de l'art en creativite et en intelligence artificielle.
Le changement de paradigme, cf TheStructureOfScientificRevolutions, propose ici implique de prendre serieusement en compte et d'appliquer les lessons de ProgrammingTheUniverse et Supersizing au domaine de la creativite de facon generale.
(definition personnelle) Creativite = action d'exploration fruit du couplage (reconfiguration du reseau de constituants) entre intention et environment
- aucunement limite a un domaine particulier
- pouvant etre dicte par la survie
- contraint par les resources disponibles, y compris pour la creation et l'utilisation d'outils
- non forcement limite a l'etre humain
clarify artificialnot mentioned in the title- support/extension/assisting vs replacement
- difficulty to produce new behaviors
- see Chaitin on math and evolution, repetitively saying he is looking for a mathematical model of creativity
- these (oui c'est possible, cf ces examples et cette theorie), antithese (non c'est impossible, cf ces problemes et resultats), synthese (ce qui a ete fait, les consequences, les tendances et que devrait apporter ces evolutions) ?
- invitation a prendre position au debut, et a la fin me dire si cela a changer dans un sens comme dans l'autre et pourquoi
- consider les outils d'Hollywood pour estimer le succes d'un scenario et d'une liste possible d'acteurs d'un film potentiel et son acceptation par le publique
- exemple de systeme de "pre-feedback" donc d'aestetique
- " A psychological explanation of creativity will be in terms of parts or processes that are themselves less-creative or in the limit not themselves creative." David Krakauer and originally from Robert Nozick, cf WithoutNotesSeptember11#UlamMemorialLectures
- Wikipedia:Abduction
and Analogy (or case-based reasoning) as studied by Douglas Hofstadter as possible paths
- cf also WithoutNotesSeptember11#LorenzoMagnani's work
- how it applies to every endeavor
- code generation (EulerGUI) and improvement (Moses, Frontier Search) also as one of the main motivation for implementation in AI
- cooking
- http://jamesoff.net/site/fun/random-recipe-generator/ or http://www.supercook.com/ (had a GA based equivalent but seems to be down for a while) with http://singularityhub.com/2009/08/03/the-robots-are-the-chefs-in-this-japanese-restaurant/
- planning
- which can be used for robotic performances and... pretty much anything
- if aesthetics remain problematic, one could consider "cheating" e.g. via AWS#MTurk
- this probably has already been done
- consider asking Person:Elise for a review
- Creativity in Science by Dean Keith Simonton, Cambridge University Press 2004
- Scientific Creativity: The Science and the Art by Dean K. Simonton, Incite Your Curiosity Lecture Series 2010
- explosion not just of actuators but also of languages to manipulate them
- simulations too, e.g. via complex systems
- even dance via some XML format
- an actuator used in science would typically be to extract information
- rather than solely produce a behavior as in art for example
- considered as a specialized epistemic actuator
- one could consider the trend of OpenCV as significant, cf own tweet
- also to consider the classical critic that the best expert of the field can do much better
- versus if the average human can not
- Will AI cause the extinction of humans? Is AI our only chance to avoid extinction?
- both ask for the same underlying question, what are "we"? which is probably the most recurrent question in culture
- thus would an AI be considered artistic when it will propose new ways to explore the question of what it is?
- yet partly already covered with self-improvement or self-optimization, e.g. Schmidtbuger and its requirement on self-models
- overall for doubts on AI, as maybe raised by Arte Tracks special Transhumanism, there is one and only one solution, learn how what you are scared about actually works
To explore
- Computers and Creativity by J. McCormack and M. d’Inverno, Springer forthcoming 2011
- actually postponed to early 2012, cf discussion with author during SonyCSL15thAnniversaryEveningSymposium
- official roadmaps? DARPA projects?
- http://fabien.benetou.fr/innovativ.it/www/wiki/ initially was the most up to date and organize personal source to exploit.
- still to exploit
- http://www.quora.com/Is-it-possible-to-study-creativity-scientifically
- http://philosophytalk.org/pastShows/Creativity.html
- http://serendip.brynmawr.edu/exchange/node/2465
- http://multiverseaccordingtoben.blogspot.com/2011/09/creatingdiscovering-new-states-of-mind.html
- The Drawbots, 2007
- discovered via art + science by Stephen Wilson, 2010
- http://www.quora.com/Is-artificial-intelligence-leading-to-artificial-creativity-If-so-what-are-some-examples-If-not-why-not
- story or script generation as used (or dream) by the video game industry
creativity
- Wikipedia:Computational creativity
- http://www.computationalcreativity.net (rather small)
- http://dl.acm.org/citation.cfm?id=581724
- http://www.thinkartificial.org/artificialcreativity/original-thoughts-intro/
- Wikipedia:Ingo Rechenberg
- http://ideac-cnrs.univ-paris1.fr
- Algorithmic aesthetics, Computer Models for Criticism and Design in the Arts by George Stiny and James Gips, University of California Press 1978
- Categories of algorithmic aesthetics by Gunter Weiss, Rutgers 2009
New proposed structure
- show the need
- personal need and how I encountered it
- generalised to most tasks required by people in the audience
- thus... why an AI would also consider it (basically Seedea:Research/Drive
)
- invite audience participation to explore the question too
- dice or audience filled seats as matrix "composer" and interpreter (cf pic2zic)
- warning, ni artistique ni scientifique mais plutot une
- demander de decrire qu'est-ce qui a ete creatif dans le processus et/ou le resultat et ce donc qui ne l'a pas ete
- possible critique: ceci n'est qu'un traitement d'information, un "mapping". Oui, mais qu'est-ce qui ne l'est pas ?
- dans cet example si cela a ete creatif, quelle ou quelles etapes l'ont ete? idee, discussions, conception, tests, execution publique, ...? (possible demande de vote a main levee)
- warning, ni artistique ni scientifique mais plutot une
- quiz
- dice or audience filled seats as matrix "composer" and interpreter (cf pic2zic)
- demonstrate via state of the art
- use last proceedings
- use layers to organise
- step back to generalise
- importance of have a proper theory
- consider defining OwnConcepts#UniversalComposer
- distinguish pattern
- importance of have a proper theory
- reconnect the need, the newly formulated theory and a praxis
- 3D printers (Wikipedia:G-code
and SKDB project)
- generalised physical effectors/actuators to robotics
- 3D printers (Wikipedia:G-code
- opening up
- future scenarii and consequences of each
- thanks
- opening Q&A
Propositions
- indexations ouvertes de repertoires problemes associes a leurs solutions
- y compris les algorithmes de parcours et les statistiques d'utilisation
- Scientometrics a la WithoutNotesApril11#MDTS mixe au KM personel ou de groupe (e.g. via son propre wiki)
- lister les repertoires (repository) de designs et inferrer la phylogeny et inferrer l'algorithme de generation (inversion) de chaque puis global (generalisation)
- ΦFP
, AlgorithmicEpistemology, ...
- mining structured repositories (scores, designs, software, ...)
- indexation du passe pour la hierarchisation et l'extraction de pattern en masse et de facon systematique, ceci meme a travers differents domaines
- multi-modal Microsoft Paint replacing the brush or the rectangle by parametrized generation based on the moment (mood, intent, ...)
- which could benefit from ExtendedLayeredModel
- maybe the synthesis of my last weeks could be that if P!=NP (which seems to be the case until proven wrong) then AI can be creative yet not necessarily significantly more efficiently than humans
- i.e. solutions exist but are costly
- also implies that niches will exist while some area will remain unchanged
Demo of own stuff to be less philosophical
- AlgorithmicEpistemology
- including periodic reminder
- can express emotion felt while writing down an interesting result, e.g. Excited
- Seedea:CombinatorialCreativity/GrammarRuleBinary
- OpenCV/ChucK project
References
- WithoutNotesOctober11#ComputationalCreativityProceedings
- Seedea:Research/Bibliography
and Seedea:Research/AutomatedArt
- http://www.routledge.com/books/details/9780805811537/
- http://www.cs.ucl.ac.uk/staff/p.bentley/evdes.html
- http://creative.canberra.edu.au/mitchell/metacreation.html
- http://www.metacreation.net (but mostly about music)
- http://art-artificial-evolution.dei.uc.pt/abstracts.htm
- http://l3dswiki.cs.colorado.edu:3232/CreativeIT/
- http://www.creativityandcognition.com/content/view/70/140/
- AI EDAM Artificial Intelligence for Engineering Design, Analysis and Manufacturing
Transition
Presentation precedente avec les automates cellulaires with Wikipedia:Rule 30 and EmergenceEnSciencesCognitives#EmergentComputationInBiochemicalNetworks. The link with the next talks will be done later.
Pourquoi
Pourquoi la creativite est-elle centrale dans la recherche artistique comme scientifique ?
- pourquoi cette problematique m'interesse-t-elle tant ?
- definition du soi (comme partage avec Paola dans le voyage a BXL)
- regarder ce que l'on a sur soit ou dans sa poche et compter le nombre d'objets uniques, cree soit-meme ou par des amis (a essayer en direct)
- experience antipub "journee sans achat"
- elargir, pourquoi cela devrait interesser le reste de la salle
- definition du soi (comme partage avec Paola dans le voyage a BXL)
- usages
- per domain, e.g. science, art, manufacturing, robotic, ...
- creativity is the root of all new solutions
- Seedea:Research/Drive
- "Pygmalion’s gift—we can imagine new worlds and intentionally make them real." (p203) TheTinkerersAccomplice#Chapter10
- tendances
- repertoires de sources mais maintenant de designs
- impression 3D, synthese ADN, assemblage atome par atome, ..?
- si LongevitityEscapeVelocity/dure de vie bcp plus longue, mieux vaut avoir une infinite d'activite a faire sinon a quoi bon
De plus, meme si le but n'est pas et ne serait jamais atteind, historiquement les outils crees pour s'en approcher deviennent fondamentaux.
Pourquoi (recherche artistique)
Peacock
- Mating Mind
- competition demande individualisation, differenciation, criteres de comparaisons sans cesse renouveles
peinture sur cave
- Etape clef dans l'histoire des explosions creatives
- quel serait l'equivalent pour une IA
- hypothese de l'outillation des animaux
- gestion des connaissances (KM)
mirroir
- Artificial Intelligence Research As Art by Stephen Wilson, 1995
- build self-model, cf autonomous robotics
- beaucoup de chercheurs en IA et sciences cognitives cherchent a se comprendre
exploration intellectuelle, etre surpris par l'execution de son propore code
- experience de pensee assistee par ordinateur (EPAO)
Pourquoi (recherche scientifique)
adam & co
Pourquoi (recherche scientifique)
paradigm shift consider own proposal of log of errors against the model, if above threshold revise the foundations, a la simulated annearling
Benefices pour une I.A.
Benefices pour une I.A. (individualisation)
- overall to
- distinguish/be unique
- personne n'est connu pour des choses dont on n'a pas l'impression qu'il les a concu
- also rational, what is needed and unique is irreplaceable, thus must be protected
- socially safe position
- distinguish/be unique
Benefices pour une I.A. (Reine Rouge)
- be more efficient
- competitively or not
- physical exploration as one of the first step, probably with genetic and physiological basis, a simple way extract information out of the environment and thus to try to establish better correlation, hopefully causation, to potentially survive (increasing CoS)
- yet maybe two sides of the same coin!
Benefices pour une I.A. (auto-amelioration)
- AI self-improvement just one boringly rational intent amongst many possible others
- already theoretically researched WithoutNotesAugust11#AFamilyOfGodelMachineImplementations, cf http://www.idsia.ch/~juergen/goedelmachine.html
- possibly WithoutNotesAugust11#Evolvability too but link to AI to clarify, surely important for ML
- already partly used in commercial products WithoutNotesJune11#CuriosityDrivenDevelopment
- already planned in AGI http://opencog.org/roadmap/
- Self-Improving Algorithms, SIAM J. Comput. April 2011
- already theoretically researched WithoutNotesAugust11#AFamilyOfGodelMachineImplementations, cf http://www.idsia.ch/~juergen/goedelmachine.html
Benefices pour une I.A. (... mais aussi)
(spec)
Mechanismes
Comment pourrait-elle acquerir des mechanismes arrivant a un tel resultat et a quel prix ?
Mechanismes (probleme de recherche)
Assimilable a un probleme de recherche dans un espace de solutions et resultant dans un resultat non encore expose, probalement donc en explorant un espace plus vaste qu'auparavant, tout du moins different.
Mechanismes (multiples contraintes)
- new in general/new only for the author, cf Boden's personal-creativity and historical-creativity (2003)
- e.g. rediscovering Newton's law by Eureqa would only belong to the 1st category
- efficiency, utility or aesthetic
- e.g. won prizes or highly ranked in a famous competition in the domain like MLcomp
- to complete with computational aesthetic material recently discovered
- theoretical/functional
- autonomous/¬autonomous e.g. interactive
- limited search space
- parametric e.g. used in architecture (Zaha Hadid) and design (WEATHER LAMP, Samuel Javelle during RADArt6)
- yet those are not "infinitely creative", it's like switching from integers to decimals as illustrated by http://xkcd.com/915/ , a classical technique of series in art
- parametric e.g. used in architecture (Zaha Hadid) and design (WEATHER LAMP, Samuel Javelle during RADArt6)
- legal
- ethical
- well known ways to express such constraints
- fitness functions (cf http://www.quora.com/Fitness-Function/Is-there-a-repository-of-fitness-functions )
- computational aesthetics
- economical utility
- spectrum rather than a binary test
- each form would thus be one of the hereafter mentioned constraint
Aesthetics and expertise
IA and creativity seems to share also the way they are perceived, i.e. the more you know how the final result was obtained, the less it appear to fit the definition.

(or two sigmoids, as done before with PersonalUX#Visualization) potentially with degrading colored from left part to right parts to represent moving away from ML to having a dedicated proper model
decreasing magic for increasing expertise but with different type of appreciation, from superficial aesthetics (which can even be learned by pigeons, cf David Krakauer Ulam second lecture thus leading to think that it could be emulated via machine learning) to deeply conceptual and its potentiality (with e.g. Kandinsky, Picasso or even history of art in general with the recent Manifesto for conceptual art, cf Art but also the study, iirc showed by Dean Keith Simonton during WithoutNotesAugust11#BeautifulMindsTheEnigmaOfGenius).
Mechanismes (multiples contraintes)
e.g. integerer le feedback de son environment, le galiriste, ses pairs, etc...
Mechanismes (systemes existants)
- General Problem Solving (mostly theoretical progresses)
- Expert systems and rules system (scalable?)
- Machine Learning (costly to train and extremely narrow)
- Evolutionary computation (cf http://www.quora.com/Why-hasnt-evolutionary-computation-been-more-successful )
- Generative art (good at abstract art and natural forms which are iterative, cf evo-devo)
- CP, POMDPs, ...
- genr8
- Agency-GP AI Lab Website A Spatial Exploration Tool for Architects Based on Evolutionary and Agent Computation
- nodal
- WithoutNotesOctober09#Adam en science
- ~re-decouverte, parametrage, ... mais pas, contrairement au titre permettant un changement de paradigme
- Emergence of Creativity: A Simulation Approach by Hrafn Th. Thórisson, Intelligent Complex Adaptive Systems, IGI Publishing 2008
- Eden by Jon McCormack, cf CSL
- creative software "star system"
- Wikipedia:Computer music#Computer-generated_scores_for_performance_by_human_players
(EMI)
- Wikipedia:AARON
- BACON.1 BACON.3* BACON.5
- Copycat by Melanie Mitchell and Douglas Hofstadter
- Metacat by James B. Marshall, 2003
- whatever opinion we might have on them, beg the question on if there is pattern of popularity
- Wikipedia:Computer music#Computer-generated_scores_for_performance_by_human_players
- AGI community discussions on creativity
- http://www.mail-archive.com/search?l=agi%40v2.listbox.com&q=creativity
- [agi] Creativity Engine Emeka Okafor, 2004
- [agi] analogy, blending, and creativity J Storrs Hall, 2007
- [agi] creativity Ben Goertzel, 2008
- [agi] [WAS The Smushaby] The Logic of Creativity Mike Tintner, 2009
- [agi] The fuzzy creative calculator. Ronald C Blue, 2009
- [agi] The Math Behind Creativity Mike Tintner, 2010
- agi@listbox.com
- Creative/Innovative Networks in AI/AGI Mike Tintner, September 2011
- com.googlegroups.opencog
- mention of Schmidhuber's work in September 2011, especially http://www.idsia.ch/~juergen/creativity.html
- according to overall mentions of creativity seems to be a "hot topic" since few weeks
- artificial-general-intelligence@googlegroups.com
- AGIRI wiki http://www.agiri.org/wiki/Special:Search?search=creativity&fulltext=Search
- http://www.mail-archive.com/search?l=agi%40v2.listbox.com&q=creativity
- BISON develop and validate a computational methodology, which facilitates bisociative information discovery in large-scale heterogeneous information environments.
- based on Wikipedia:The Act of Creation
with the concept of bisociation
- seems to have no software produced nor paper published since 2010
- based on Wikipedia:The Act of Creation
Et probleme recurrent du jugement du resultat, i.e. Wikipedia:Demarcation problem en allant plus loin que le deja difficile a etablir critere de nouveaute, cf Wikipedia:Prior art
. Doute par exemple d'Alain Connes et Stanislas Dehaene sur la difficulte voir l'impossible pour le moment de criteres aestetiques (cf les differentes pistes de solutions de computational aesthetics trouvees recemment).
Mechanismes (synthese)
|
Gather context, eventually incrementally precisely as it can not be expected to be perfect, and intent. |
⇒ |
Shortest description that can maximize expressive power by leveraging
Step by step process
|
⇒ |
Maximizing aesthetics/utility/efficiency while minimizing predictability by potential observers. Note that the intent and the observers can be equivalent but do not have to, the iteration of the process can make them match. |
cf OwnConcepts#UniversalComposer
Mechanismes (complexite)
- computational complexity of creativity
- Complexity to Creativity Table of Contents by Ben Goertzel, Plenum Press 1997
- WithoutNotesSeptember11#AviWigderson Knowledge, Creativity and P versus NP
- cf previously watched but mot recent IAS lecture WithoutNotesJanuary10#AviWigderson with a definition of P/NP related to creativity
- related to search space? AIT?
- WithoutNotesSeptember11#ScottAaronsonCMU2011BuhlLecture and his No Super-Search Principle
- applied to finance in WithoutNotesSeptember11#PhilipMaymin
- if one solves P = NP one solves creativity
- but if one automates creativity efficiently one can solve P vs NP
- section 6.3. Is Mathematics "Creative"? of K. Podnieks Consequences of Incompleteness Theorems
Consequences
Si de tels mechanismes etaient efficacement implementes dans une machine, quelles seraient les consequences sur le plan intellectuel, social et humain ?
Consequences (intellectuelles)
Cf Health#ImpactOfInformationTechnologyOnTheBrain et WithoutNotesSeptember11#TheNeuronalRecyclingHypothesis
Consequences (sociales)
- authorship, law and protection
- copyright on author? production?
- current examples
- Media Lab logo
- Weather lamp, cf RADArt6
- Genie in the Machine
- Computers and Creativity by Robert Plotkin, Infobase October 2011
- but actually a deeper question on complexity
- basically if this if artificial creativity is possible it implies a different structure of complexity that what was currently thought of hitherto
- that reaching an efficiently solution might not require more than copying an existing one, consequently having a very large compensation for searching for it made sense only until a generalist solution was found
- thus if payback provided by e.g. patents is based on risk, the system has be adapted
- with a generalist solution risk approaches 0 thus compensation should become lower too
- basically if this if artificial creativity is possible it implies a different structure of complexity that what was currently thought of hitherto
- WithoutNotesFebruary11#MartinFord
Consequences (humain)
- Neme revolution qui bouscule l'humain du centre du "monde".
- de nouveau devoir se redefinir et chercher son cote unique
Conclusion
- even if creativity can not be fully automatized (because of technical limits, mainly of high complexity thus inefficiency, or societal limits, basically of adoption and acceptation) change through assistance and support will radically alter the landscape of what can be done
- yet, this does not have to be felt as a revolution for those who keep track of the possibilities, what has already been done and what is most likely to change soon
- as defined in "A sudden, vast change in a situation, a discipline, or the way of thinking and behaving." cf https://secure.wikimedia.org/wiktionary/en/wiki/revolution
- as illustrated by http://www.paleofuture.com/blog/2010/2/17/robots-will-kill-music-1930.html revolutions are mainly a question of perspective, if you leverage a change its a transition, if not you get revolutionized
- yet, this does not have to be felt as a revolution for those who keep track of the possibilities, what has already been done and what is most likely to change soon
- si tous ces aspets positifs et cette possible fesabilite technique
- pourquoi cela ne fonctionne pas deja ?
- pas forcement visible
- effet de percolation, tant qu'il n'y a pas un degree suffisant d'interconnection, peut de resultat mais pourtant beaucoup d'efforts
- prediction d'une explosion creative sous peu ?
- biais social
- sensation de perte de controle par rapport aux outils utilises
- vouloir un produit "artisanal"
- toujours plus couteux que de demander a un specialiste
- voir aussi http://www.no-free-lunch.org
- pourquoi cela ne fonctionne pas deja ?
[Post-presentation reflection
- apply creativity to generate public policies
Peut-on rendre une I.A. créative ?
Creativite artificielle,
Generer et utiliser de la nouveaute
avec une regle d'a...... c......... ?
Plan
- Pourquoi la creativite est-elle centrale dans la recherche artistique comme scientifique ?
- Quels benefices pour une I.A. ?
- Comment pourrait-elle acquerir des mechanismes arrivant a un tel resultat et a quel prix ?
- Si de tels mechanismes etaient efficacement implementes dans une machine, quelles seraient les consequences sur le plan intellectuel, social et humain ?
Pourquoi
Pourquoi la creativite est-elle centrale dans la recherche artistique comme scientifique ?
Pourquoi (recherche artistique)
Pourquoi (recherche artistique)

Cave de Chauvet, 30 000 av. J.C.
Pourquoi (recherche artistique)
Pourquoi (recherche artistique)
Mais surtout ...
Fra Mauro map

Here be dragons.
Pourquoi (recherche scientifique)


Pourquoi (recherche scientifique)
mais ... comment Kuhn qualifirait cela ?
Pourquoi (recherche scientifique)
- "Allegedly, the scientific method encapsulates the rules of scientific rationality. It may be that those rules could not account for the creative side of science — the generation of new hypotheses." (own emphasis) http://plato.stanford.edu/entries/thomas-kuhn/
- "When enough significant anomalies have accrued against a current paradigm, the scientific discipline is thrown into a state of crisis, according to Kuhn. During this crisis, new ideas, perhaps ones previously discarded, are tried." (own emphasis)Wikipedia:Paradigm shift
Donc pourquoi s'interesser a la creative ?
"Il ne nous reste exactement que l'inventivite.[...] Aujourd'hui le travail intellectuel est oblige d'etre un travail intelligent et non un travail repetitif comme il a ete jusqu'a maintenant."
Michel Serres, INRIA 2007

Benefices pour une I.A.
En quoi une intelligence artificielle beneficierait-elle a etre creative ?
Benefices pour une I.A. (Reine Rouge)

Benefices pour une I.A. (auto-amelioration)

mais...
Benefices pour une I.A. (individualisation)

Benefices pour une I.A. (... mais aussi)
C'est dans les specs.
- #7 Randomness and Creativity, 1956 Dartmouth Proposal for Artificial Intelligence
- Creativity and artificial intelligence by Margaret A. Boden, 1998 Artificial Intelligence
- Computer Models of Creativity by Margaret A. Boden, 2009 AI Magazine
Mechanismes
Comment pourrait-elle acquerir des mechanismes arrivant a un tel resultat et a quel prix ?
Mechanismes (probleme de recherche)
PS : recherche au sense algorithmique, "search" en anglais
Mechanismes (multiples contraintes)

Mechanismes (multiples contraintes)
... la reponse a la question initiale "Peut-on rendre une I.A. creative ?" depend donc de ces criteres ...
Mechanismes (multiples contraintes)
... mais attention a ce que ceux-ci soient aussi appliques aux artistes et chercheurs (qui ne sont pas des I.A.) !
Mechanismes (systemes existants)
- correlation, statistiques, apprentissage machine (ML), intelligence artificielle, intelligence artificielle de niveau humain (HLAI), intelligence artificielle generaliste (AGI), intelligence artificielle universelle (UAI)
genr8
, nodal
, Adam
, Eve, Eden, EMI, Jade What kind of murderer has fiber?, AARON
, BACON.*, Eureqa
Copycat
, Metacat
, The Painting Fool
, ~OpenCog, BISON, ...
Mechanismes (synthese)


Mechanismes (complexite)

Paul Erdos dans N is a Number (extrait)
Consequences
Si de tels mechanismes etaient efficacement implementes dans une machine, quelles seraient les consequences sur le plan intellectuel, social et humain ?
Consequences (intellectuelles)

Beth Rekow
Consequences (sociales)

Consequences (humain)

...Conclusion
- le probleme n'est pas resolu a l'heure actuelle
- mais de serieux progres dans tous les domaines etudies
- peinture, musique, humour, scenario de jeux ou films, science, ...
- aesthetique
- des domaines connexes
- sciences cognitive avec l'enaction et les boucles sensorimotrices
- repertoires de designs e.g. OHL
- robotique en particuliar au niveau de la vision avec OpenCV, ...
- mais de serieux progres dans tous les domaines etudies
Tout ceci laisse entrevoir une unification possible non seulement au niveau du modele et donc se son implementation mais aussi d'une applicabilite tres large de part l'aspect general des languages de programmation et des actuateurs associes.
Laboratoires dedies
- Creative Machines at Cornell
- discovered before with Eureqa in ElegantSolutions#ExperimentsTools but then named Cornell Computational Synthesis lab and again few days ago with EndlessForms
- Computational Creativity Group at Imperial College LondonCreative Systems Lab at the University of Sussex
- Computational Creativity at Nicolaus Copernicus University
- Creative Systems at University of Coimbra
- including Francisco Câmara, author of Creativity and Artificial Intelligence: A Conceptual Blending Approach 2007 book and the Divago system
Conferences
- International Joint Workshop on Computational Creativity (IJWCC)
- International conference on computational creativity (ICCC)
Avec leurs publications en acces libre.
Evenements
- Sony CSL The Performing Sciences
vendredi dernier
- Stanford AI-Class / ML-Class
- dans ~1 semaine
- groupes d'etudes a Paris
- l'alternative "other-ai.org"
- PhiloWeb 2011
- mercredi a Salonika en Grece
- A Critique of Pure AI: Comparing the Social Foundations of Artificial Intelligence and the Web by Harry Halpin and Jacob Blumenthal
- Mathématiques, un dépaysement soudain a la Fondation Cartier pour l’art contemporain
- 21 octobre 2011 au 19 février 2012
CONTENT
CONTACT
UPDATES
LAST TWEET







RSS for this page only

