Page automatisée à partir de multiples flux RSS / Atom
Catalogues d’articles scientifiques
SAGE Publications: Collective Intelligence: Table of Contents Table of Contents for Collective Intelligence. List of articles from both the latest and ahead of print issues.
- Synch.Live: Collective problem-solving through flocking motion associated with higher connectedness to others
par Madalina I. Sas, Pedro A.M. Mediano, Fernando E. Rosas, Hillary Leone, Andrei Sas, Christopher Lockwood, Henrik J. Jensen, Daniel Bor14615Centre for Complexity Science, Imperial College London, London, UK2Department of Computing, Imperial College London, London, UK3Division of Psychology and Language Sciences, University College London, London, UK4Centre for Psychedelic Research, Department of Brain Sciences, Imperial College London, London, UK51948Sussex AI and Centre for Consciousness Science, Department of Informatics, University of Sussex6Centre for Eudaimonia and Human Flourishing, University of Oxford, Oxford, UK7Synch.Live, Independent Artist, New York, USA8Independent Artist, London, UK9Independent Researcher, Luxembourg, Luxembourg10Department of Mathematics, Imperial College London, London, UK114617Department of Psychology, Queen Mary University of London, London, UK1298528Department of Psychology, University of Cambridge, Cambridge, UK1398528Centre for Brain and Behaviour, School of Biological and Behavioural Science, Queen Mary University of London, London, UK le 2026-04-22 à 5:18 AMCollective Intelligence, Volume 5, Issue 2, April-June 2026. <br/>Background:Collective self-organising behaviour is ubiquitous in nature, whereby complex patterns emerge from the local interactions between individuals. Yet in humans, most group behaviour is attributed to explicit central control or pre-planning, …
- Incorporating memory into bounded confidence models of probabilistic social learningpar Jonathan Lawry1School of Engineering Mathematics and Technology, 1980University of Bristol, Bristol, UK le 2026-04-02 à 3:51 AM
Collective Intelligence, Volume 5, Issue 2, April-June 2026. <br/>In social learning models, truth-seeking agents learn both individually from direct evidence and socially by pooling beliefs with others. That learning can be undermined by two types of unreliable agents: zealots, who do not learn and promote the same …
- World-Making:How individual social intelligence leads to collective transformational wisdom.par RSS-Bridge le 2026-06-01 à 12:00 AM
Psychology meaningfully contributes to the creation of future societies. We first discuss the theory underlying the world-making potential of psychology. We outline the importance of social intelligence as a central process for transformational wisdom to emerge collectively and thereby contribute to the formation of future societies. Next, we apply this theoretical approach to immigration, showing how transformational wisdom can happen through the development of social intelligence used to navigate complex multicultural societies. We use two ethnographic case studies – in Ireland and Denmark – to illustrate how transformational wisdom is collectively achieved, and when scaled up, can lead to forming more tolerant and harmonious Western liberal democracies.
- Les apports des intelligences artificielle et collective dans la construction d’un outil de ludopédagogiepar RSS-Bridge le 2026-05-07 à 12:00 AM
International audience
- The Intelligence That Was Never Artificial: LLMs as Collective Human Cognition and the Cybernetics That Predicted Thempar RSS-Bridge le 2026-04-18 à 12:00 AM
v3.1 (Restored v3 content with corrected PDF typography (LaTeX via pandoc+pdfTeX), dual affiliation (Ronin Institute + IamI.Earth Foundation), ORCID 0009-0004-2876-0025, and both contact emails on cover. Supersedes v4/v5/v6 (deleted).)In 1907, Francis Galton demonstrated that the median estimate of 787 people guessing an ox’s weight was accurate to within 0.8 percent, outperforming any individual expert (Galton, 1907). We argue that LLM capabilities are best understood as structured aggregation of collective human intelligence, not autonomous machine reasoning. Transformer architectures provide the organizing substrate through which this aggregation becomes computationally tractable: necessary infrastructure, but not an independent source of intelligence. The semantic content of what LLMs know, the facts they retrieve, the reasoning patterns they deploy, and the linguistic competence they display, originates in the training corpus. The architecture provides the syntactic engine that compresses and recombines this collective knowledge. If this analysis is correct, the intelligence was never artificial; it was human intelligence, computationally reorganized. We trace the historical erasure of this insight. Wiener’s cybernetics (1948) described intelligence as emergent from feedback systems. The Dartmouth conference (1956) reframed this relational phenomenon as “Artificial Intelligence,” severing the intelligence from its collective human source. We show that three terms central to modern AI discourse obscure the technology’s actual mechanism: “Artificial” creates otherness and enables ownership; “Intelligence” misattributes agency to the product; “Training” disguises extraction of humanity’s intellectual commons as pedagogy. We ground this claim through proven mathematical identities (cross-entropy pretraining implements the linear opinion pool; RLHF implements Borda count and logarithmic opinion pooling), the Diversity Prediction Theorem, and the Conditional Jury Theorem, arguing that LLM training is best understood as high-dimensional judgment aggregation. The framework generates predictions that scaling laws cannot make, including diversity-disproportionality, and provides parsimonious retrodictions of known phenomena including tail-first model collapse under independence violation. We present an evidence synthesis drawing on Liu et al.’s (2024) 512 controlled training runs and converging results from four independent lines of research, supporting the diversity prediction, and conclude that the naming determines who benefits from collective human intelligence computationally reorganized. v5 (2026-04-17): PDF rebuilt via pandoc + pdfTeX toolchain. v4 used a non-LaTeX PDF library that degraded typography and changed the cover page; v5 restores the proper paper template. Content is unchanged from v4. Author metadata updated to dual affiliation (Ronin Institute and IamI.Earth Foundation) and the nyx.redondo@ronininstitute.org email. v6 (2026-04-17): Cover page updated to include the nyx@iami.earth contact email alongside nyx.redondo@ronininstitute.org. No scientific content or typography changed from v5.
- The Intelligence That Was Never Artificial: LLMs as Collective Human Cognition and the Cybernetics That Predicted Thempar RSS-Bridge le 2026-04-18 à 12:00 AM
v3.1 (Restored v3 content with corrected PDF typography (LaTeX via pandoc+pdfTeX), dual affiliation (Ronin Institute + IamI.Earth Foundation), ORCID 0009-0004-2876-0025, and both contact emails on cover. Supersedes v4/v5/v6 (deleted).)In 1907, Francis Galton demonstrated that the median estimate of 787 people guessing an ox’s weight was accurate to within 0.8 percent, outperforming any individual expert (Galton, 1907). We argue that LLM capabilities are best understood as structured aggregation of collective human intelligence, not autonomous machine reasoning. Transformer architectures provide the organizing substrate through which this aggregation becomes computationally tractable: necessary infrastructure, but not an independent source of intelligence. The semantic content of what LLMs know, the facts they retrieve, the reasoning patterns they deploy, and the linguistic competence they display, originates in the training corpus. The architecture provides the syntactic engine that compresses and recombines this collective knowledge. If this analysis is correct, the intelligence was never artificial; it was human intelligence, computationally reorganized. We trace the historical erasure of this insight. Wiener’s cybernetics (1948) described intelligence as emergent from feedback systems. The Dartmouth conference (1956) reframed this relational phenomenon as “Artificial Intelligence,” severing the intelligence from its collective human source. We show that three terms central to modern AI discourse obscure the technology’s actual mechanism: “Artificial” creates otherness and enables ownership; “Intelligence” misattributes agency to the product; “Training” disguises extraction of humanity’s intellectual commons as pedagogy. We ground this claim through proven mathematical identities (cross-entropy pretraining implements the linear opinion pool; RLHF implements Borda count and logarithmic opinion pooling), the Diversity Prediction Theorem, and the Conditional Jury Theorem, arguing that LLM training is best understood as high-dimensional judgment aggregation. The framework generates predictions that scaling laws cannot make, including diversity-disproportionality, and provides parsimonious retrodictions of known phenomena including tail-first model collapse under independence violation. We present an evidence synthesis drawing on Liu et al.’s (2024) 512 controlled training runs and converging results from four independent lines of research, supporting the diversity prediction, and conclude that the naming determines who benefits from collective human intelligence computationally reorganized. v5 (2026-04-17): PDF rebuilt via pandoc + pdfTeX toolchain. v4 used a non-LaTeX PDF library that degraded typography and changed the cover page; v5 restores the proper paper template. Content is unchanged from v4. Author metadata updated to dual affiliation (Ronin Institute and IamI.Earth Foundation) and the nyx.redondo@ronininstitute.org email. v6 (2026-04-17): Cover page updated to include the nyx@iami.earth contact email alongside nyx.redondo@ronininstitute.org. No scientific content or typography changed from v5.
- The Intelligence That Was Never Artificial: LLMs as Collective Human Cognition and the Cybernetics That Predicted Thempar RSS-Bridge le 2026-04-17 à 12:00 AM
In 1907, Francis Galton demonstrated that the median estimate of 787 people guessing an ox’s weight was accurate to within 0.8 percent, outperforming any individual expert (Galton, 1907). We argue that LLM capabilities are best understood as structured aggregation of collective human intelligence, not autonomous machine reasoning. Transformer architectures provide the organizing substrate through which this aggregation becomes computationally tractable: necessary infrastructure, but not an independent source of intelligence. The semantic content of what LLMs know, the facts they retrieve, the reasoning patterns they deploy, and the linguistic competence they display, originates in the training corpus. The architecture provides the syntactic engine that compresses and recombines this collective knowledge. If this analysis is correct, the intelligence was never artificial; it was human intelligence, computationally reorganized. We trace the historical erasure of this insight. Wiener’s cybernetics (1948) described intelligence as emergent from feedback systems. The Dartmouth conference (1956) reframed this relational phenomenon as “Artificial Intelligence,” severing the intelligence from its collective human source. We show that three terms central to modern AI discourse obscure the technology’s actual mechanism: “Artificial” creates otherness and enables ownership; “Intelligence” misattributes agency to the product; “Training” disguises extraction of humanity’s intellectual commons as pedagogy. We ground this claim through proven mathematical identities (cross-entropy pretraining implements the linear opinion pool; RLHF implements Borda count and logarithmic opinion pooling), the Diversity Prediction Theorem, and the Conditional Jury Theorem, arguing that LLM training is best understood as high-dimensional judgment aggregation. The framework generates predictions that scaling laws cannot make, including diversity-disproportionality, and provides parsimonious retrodictions of known phenomena including tail-first model collapse under independence violation. We present an evidence synthesis drawing on Liu et al.’s (2024) 512 controlled training runs and converging results from four independent lines of research, supporting the diversity prediction, and conclude that the naming determines who benefits from collective human intelligence computationally reorganized. v5 (2026-04-17): PDF rebuilt via pandoc + pdfTeX toolchain. v4 used a non-LaTeX PDF library that degraded typography and changed the cover page; v5 restores the proper paper template. Content is unchanged from v4. Author metadata updated to dual affiliation (Ronin Institute and IamI.Earth Foundation) and the nyx.redondo@ronininstitute.org email. v6 (2026-04-17): Cover page updated to include the nyx@iami.earth contact email alongside nyx.redondo@ronininstitute.org. No scientific content or typography changed from v5.
Laboratoires
- Alice Cai, Iman YeckehZaare, Shuo Sun, Vasiliki Charisi, Xinru Wang, Aiman Imran, Robert Laubacher, Alok Prakash, Thomas W. Malone, Where can AI be us…par RSS-Bridge le 2026-04-19 à 3:00 AM
Alice Cai, Iman YeckehZaare, Shuo Sun, Vasiliki Charisi, Xinru Wang, Aiman Imran, Robert Laubacher, Alok Prakash, Thomas W. Malone, Where can AI be used? Insights from a deep ontology of work activities, CCI working paper, arXiv, Mar 2026.
- Bridge returned error 500! (20564)par RSS-Bridge le 2026-04-21 à 3:29 AM
Details Type: Exception Code: 500 Message: Could not request AI4CI Outputs via proxy. HTTP Code: 0 File: lib/utils.php Line: 252 Trace #0 index.php(73): RssBridge->main() #1 lib/RssBridge.php(39): RssBridge->{closure}() #2 lib/RssBridge.php(37): CacheMiddleware->__invoke() #3 middlewares/CacheMiddleware.php(44): RssBridge->{closure}() #4 lib/RssBridge.php(37): ExceptionMiddleware->__invoke() #5 middlewares/ExceptionMiddleware.php(17): RssBridge->{closure}() #6 lib/RssBridge.php(37): SecurityMiddleware->__invoke() #7 middlewares/SecurityMiddleware.php(19): RssBridge->{closure}() #8 lib/RssBridge.php(37): MaintenanceMiddleware->__invoke() #9 middlewares/MaintenanceMiddleware.php(10): RssBridge->{closure}() #10 lib/RssBridge.php(37): BasicAuthMiddleware->__invoke() #11 middlewares/BasicAuthMiddleware.php(13): RssBridge->{closure}() #12 lib/RssBridge.php(37): TokenAuthenticationMiddleware->__invoke() #13 middlewares/TokenAuthenticationMiddleware.php(10): RssBridge->{closure}() #14 lib/RssBridge.php(34): DisplayAction->__invoke() #15 actions/DisplayAction.php(54): DisplayAction->createResponse() #16 actions/DisplayAction.php(89): Ai4ciOutputsBridge->collectData() #17 bridges/Ai4ciOutputsBridge.php(35): returnServerError() #18 lib/utils.php(252) Context Query: action=display&bridge=Ai4ciOutputsBridge&format=Atom Version: 2025-01-26 OS: Linux PHP: 8.3.23 Go back Find similar bugs Create GitHub Issue https://all-as.one
Vidéos
- On visite l’expo FOULES, avec @Lea_Bello et @CyrusNorth
par Fouloscopie le 2026-04-15 à 12:37 AML’exposition “FOULES” est prolongée jusqu’au 26 MAI 2024, à la Cité des Sciences et de l’Industrie, à Paris. Si vous l’avez déjà visitée, n’hésitez pas à me faire un retour en commentaire. Bonne visite !Et ne manquez pas la chaîne youtube de Léa : https://www.youtube.com/@UC-jcC3zw4pXOn8t3keKY__g Et celle de Cyrus : https://www.youtube.com/@UCah8C0gmLkdtvsy0b2jrjrw *** POUR ME SOUTENIR ***➡️ sur Tipeee : https://fr.tipeee.com/fouloscopie➡️ sur Utip : https://utip.io/fouloscopie Merci !*** MES RECHERCHES ***Mon laboratoire de recherche à l’institut Max Planck : https://www.mpib-berlin.mpg.de/research/research-centers/adaptive-rationalityMa thèse de doctorat : http://mehdimoussaid.com/TheseMoussaid.pdf
- Combien faut-il de personnes pour lancer UNE RÉVOLUTION ? ✊✊✊
par Fouloscopie le 2026-04-12 à 2:28 AMGrande nouvelle : je suis en train de préparer la deuxième saison des expériences participatives ! Inscrivez-vous ici pour être informé du lancement de la campagne, soutenir le projet et participer aux expériences : ➡️ https://www.kisskissbankbank.com/fr/projects/fouloscopie-100x-saison2 ⬅️Rendez-vous à la rentrée pour le démarrage du projet !Bonne vacances !
- Les bousculades dans les stades de foot
par Fouloscopie le 2026-04-11 à 7:27 PMErratum: la pression se mesure effectivement en Newton par mètre carré. Désolé pour l’imprécision.Pour soutenir la chaîne :➡️ sur Tipeee : https://fr.tipeee.com/fouloscopie➡️ sur KissKiss : https://www.kisskissbankbank.com/fr/projects/fouloscopieMerci à mon invité Pascal Viot, pour en savoir plus sur ses activités :➡️ https://www.issue.ch/Pour en savoir plus sur mes recherches :➡️ Mon laboratoire de recherche à l’institut Max Planck : https://www.mpib-berlin.mpg.de/research/research-centers/adaptive-rationality➡️ Ma thèse de doctorat : http://mehdimoussaid.com/TheseMoussaid.pdfEnfin quelques références pour la vidéo :➡️ Pour l’étude des pressions physiquesWang, C., Shen, L., & Weng, W. (2020). Experimental study on individual risk in crowds based on exerted force and human perceptions. Ergonomics, 63(7), 789-803.➡️ Pour un rapport complet sur le drame de Hillsborough :Nicholson, C. E., & Roebuck, B. (1995). The investigation of the Hillsborough disaster by the Health and Safety Executive. Safety Science, 18(4), 249-259.➡️ Le blog de Keith Still donne des valeurs de références pour la pression: https://www.gkstill.com/CV/Modelling/Pressure.html➡️ Et le lien vers la vidéo du journal Le Monde : https://youtu.be/6_o8EwK-m7o
- La science des foules en conditions EXTRÊMES
par Fouloscopie le 2026-04-08 à 4:36 AMQuelques liens utiles : Mon dernier livre, “A-t-on besoin d’un chef ? “👉 https://allary-editions.fr/products/mehdi-moussaid-a-t-on-besoin-dun-chef?srsltid=AfmBOoq8ke_n_HgCYM3mOLzXknt-DwMOSiDWOOnOsTZO5MY_n3Fi_zxjLa terre au carré, sur France Inter : 👉 https://youtu.be/2dw9FSgI138?si=__HsT6Msa_IKBd0_Mon passage sur l’excellent podcast de Patrick Baud :👉 https://open.spotify.com/episode/2y61v5ihiuR80KSHDS0JWa?si=WYg5Wip9RiqjEws_3j7wNQPour vous inscrire à l’avant-première du 27 avril : 👉 https://clubdeletoile.fr/programmation/le-dilemme-moral-fouloscopie/Pour soutenir la chaîne :➡️ sur Tipeee : https://fr.tipeee.com/fouloscopie➡️ sur KissKiss : https://www.kisskissbankbank.com/fr/projects/fouloscopieEnfin quelques références bibliographiques :** L’article fondateur de John Fruin sur le diagramme fondamental :Fruin, J. (1970). Designing for pedestrians a level of service concept. Polytechnic University.** Les premières analyse de San Fermin par Iker Zuriguel : Parisi et al. (2021). Pedestrian dynamics at the running of the bulls evidence an inaccessible region in the fundamental diagram. PNAS, 118(50)** L’article récent de Denis Bartolo sur les vortex de foules :Gu et al. (2025). Emergence of collective oscillations in massive human crowds. Nature, 638(8049), 112-119.
- Les 3 lois de L’ATTRACTION SOCIALE
par Fouloscopie le 2026-02-10 à 10:25 PMVous pouvez lire gratuitement le premier chapitre de mon livre sur le site de l’éditeur :👉 https://allary-editions.fr/products/mehdi-moussaid-a-t-on-besoin-dun-chef?srsltid=AfmBOoq8ke_n_HgCYM3mOLzXknt-DwMOSiDWOOnOsTZO5MY_n3Fi_zxjPour soutenir la chaîne :➡️ sur Tipeee : https://fr.tipeee.com/fouloscopie➡️ sur KissKiss : https://www.kisskissbankbank.com/fr/projects/fouloscopieEnfin quelques références bibliographiques :** Le fameux « Karaté-club » de Wayne Zackary : Zachary, W. W. (1977). An information flow model for conflict and fission in small groups. Journal of anthropological research, 33(4), 452-473.** Les premieres analyses mathématiques des réseaux aléatoires par Erdos et Renyi :Erdős, P.; Rényi, A. (1959). “On Random Graphs. I” (PDF). Publicationes Mathematicae. 6 (3–4): 290–297.** Les sociogrammes de Jacob Moreno : Moreno, J. L. (1934). Who shall survive?: A new approach to the problem of human interrelations.(L’ouvrage est en version complète ici : https://archive.org/details/whoshallsurviven00jlmo/page/4/mode/2up)** L’étude de Jure Leskovec sur l’évolution des réseaux sociaux en ligne : Leskovec, J., Backstrom, L., Kumar, R., & Tomkins, A. (2008, August). Microscopic evolution of social networks. In Proceedings of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining (pp. 462-470).** La découverte de l’attachement préférentiel par Albert Barabasi :Barabási, A. L., & Albert, R. (1999). Emergence of scaling in random networks. science, 286(5439), 509-512.