Author: cch_admin

UM6P Science Week, 21 February 2023, 52mn

Introductory texts (English, French and Arabic) by Hind Aboulazm 
Drawing by Sophie Lenormand

Luc Steels is a linguist by training and one of the main thinkers on artificial intelligence and robotics. He offers  alternative paradigms to artificial intelligence, addressing issues about body model, language, origins of language, origins of symbols, evolution, cooperation, and about how a mind could grow in artificial agents.

AI systems are being utilized to tackle complex problems through (deep) reinforcement learning. However, their limitations are becoming evident, particularly in areas like perception, where sensitivity to specific features can make them fragile. For instance, AI systems struggle to generalize when faced with different lighting conditions in image recognition tasks.

Recently, the big talk about conversational interfaces has gained significant attention, with systems like Google Lambda and OpenAI’s GPT, valued at billions of dollars. However, there are concerns about the accuracy and trustworthiness of the generated content. Misinformation and errors in texts produced by conversational AI pose risks to the information ecosystem.

Another issue lies in the use of AI to generate new content, such as images and code. Copyright and accuracy problems arise when AI utilizes existing works to create new ones. The inherent flaws in AI technology make it challenging to ensure the correctness and reliability of generated content.

Luc Steels advocates for combining a knowledge-based approach with data-driven AI to address the complexity of real-world problems. He emphasizes the importance of considering fundamental problems before releasing AI technologies.

In his work, Steels draws from the perspective of complexity, tracing its roots in AI research since the 1950s. He explores the idea of “scala naturae” which focuses on intentional stances and how entities possess self-recognition, subjective experience, and awareness.

Steels conducts numerous experiments to understand how complexity emerges in artificial systems. By studying the interactions between multiple autonomous robots and their emergent behaviors, he gains insights into cognition and mental capabilities. This complex systems approach helps shed light on open problems in neuroscience, such as the origins of color categories and self-awareness in robots.

In conclusion, Luc Steels is a pioneering linguist and AI researcher who seeks alternative paradigms for artificial intelligence.His work emphasizes the need to address fundamental problems and complexities in AI technologies to ensure their responsible and effective use in real-world applications.

Hind Aboulazm

_________________________________

Luc Steels est un linguiste et l’un des principaux penseurs en intelligence artificielle et en robotique. Son travail se concentre sur le développement de paradigmes alternatifs pour l’intelligence artificielle, abordant des questions telles que le modèle corporel, le langage, les origines du langage, les symboles, l’évolution, la coopération et le développement de l’esprit chez les agents artificiels.

Il souligne les limites de l’intelligence artificielle actuelle, notamment dans le domaine de la perception et de la fragilité des systèmes face à la généralisation des caractéristiques. Les interfaces conversationnelles suscitent également des inquiétudes concernant la fiabilité et l’exactitude du contenu généré.

Steels préconise une approche combinant l’intelligence artificielle basée sur la connaissance et celle axée sur les données pour relever les défis complexes du monde réel. Il s’appuie sur la perspective de la complexité, explorant les positions intentionnelles et la conscience chez les entités.

Ses expériences sur les systèmes complexes et les comportements émergents des robots autonomes enrichissent notre compréhension de la cognition et des capacités mentales. Elles ouvrent de nouvelles perspectives en neurosciences, notamment sur les origines des catégories de couleurs et la conscience de soi chez les robots.

En résumé, Luc Steels est un pionnier en linguistique et intelligence artificielle, dont les recherches contribuent à repenser les approches de l’IA et à promouvoir une utilisation responsable et efficace des technologies dans le monde réel.

H.A.

Luc Steels, scientifique et artiste belge né en 1952, pionnier de l’intelligence artificielle en Europe, est connu pour ses travauxsur les systèmes experts, la robotique comportementale, la vie artificielle et la linguistique computationnelle évolutive. Membre de l’ICREA à Barcelone, il a été directeur fondateur du Laboratoire d’IA de la VUB et du Sony Computer Science Laboratory à Paris.  

Luc Steels a dirigé de nombreux ouvrages dont la publication de The Artificial Life Route to Artificial Intelligence. Building Situated Embodied Agents (avec Rodney Brooks, New Haven, Lawrence Erlbaum Associates, 1993), The Biology and Technology of Intelligent Autonomous Agents (Berlin, Springer-Verlag, 1995) et Cooperative Robots : A Case Study in Animal Robotics (avec D. McFarland, Cambridge, Ma, MIT Press, 1997).

Source: https://fr.wikipedia.org/wiki/Luc_Steels

UM6P Science Week, 21 February 2023, 1h26mn
Talk followed by a roundtable with Stephen Wolfram, Hervé Zwirn and David Chavalarias
Moderator: Fouad Laroui

Introductory texts (English, French and Arabic) by Hind Aboulazm
Drawing by Sophie Lenormand

In his talk, Gregory Chaitin delves into the intricate realm of metamathematics, contemplating its role as both a companion and adversary to the world of mathematics. Metamathematics, an endeavor aimed at introspectively scrutinizing mathematics through mathematical means, traces its lineage back to Leibniz, its grandparent, and its parental figure, David Hilbert. At its core, this discipline relies upon formal axiomatic systems — frozen versions of mathematics viewed from a vantage point above. Such systems serve as the crucible for reasoning and mathematical proof, rigorously delineated to exclude any subjective elements through the apparatus of symbolic logic. Chaitin sketches the historical underpinnings of metamathematics, with Hilbert’s assertion that mathematics is inherently binary — black or white — when executed within a formal axiomatic system. This approach necessitates defining and formalizing the system’s structure, exemplified by Alfred Tarski’s work on geometry. The Peano arithmetic system and the Zermelo-Fraenkel set theory stand as emblematic formal axiomatic frameworks.

Intriguingly, Chaitin introduces the concept of an “elegant program” —  minimalistic code that yields precise output — providing a gateway to the incompleteness theorem. He expounds on the challenges in proving an elegant program, accentuating the limitations of formal axiomatic systems in achieving such a task. The notion of incompleteness, which perturbs the ambition of a universal mathematical framework, becomes a recurrent motif throughout the presentation. Chaitin’s intellectual trajectory culminates in his assertion that any formal axiomatic system —  no matter its intricacy — can only grasp an infinitesimal fraction of the boundless intricacies intrinsic to the Platonic realm of pure mathematics. He underscores the necessity of expanding axiomatic horizons and embracing new paradigms to engage with the enigma of undecidability and foster further mathematical progress. While complex and often shrouded in philosophical implications, this endeavor offers a richer perspective into the evolving nature of mathematics and its interactions with the human intellect.

The discussion with Stephen Wolfram, Hervé Zwirn and David Chavalarias that followed the presentation witnessed a synthesis of viewpoints from various luminaries, oscillating between optimism and caution regarding the future of mathematics. Contemplations ranged from the compatibility of mathematics with science despite the specter of incompleteness to the intrinsic value of computational irreducibility in the fabric of existence. Chaitin’s exposition ultimately illuminated the profound interplay between mathematics, cognition, and the boundless tapestry of human inquiry.

Hind Aboulazm

Gregory Chaitin, born in 1947, is an Argentine-American mathematician and computer scientist. From the late 1960s onward, he has made significant contributions to algorithmic information theory and metamathematics, including a notable result in computer theory analogous to Gödel’s incompleteness theorem. Chaitin, along with Andrei Kolmogorov and Ray Solomonoff, is viewed as one of the pioneers of the algorithmic complexity field, often referred to as Solomonoff-Kolmogorov-Chaitin complexity or program-size complexity. The emergence of algorithmic information theory as a core aspect of theoretical computer science, information theory, and mathematical logic owes much to the contributions of figures like Solomonoff, Kolmogorov, Martin-Löf, and Leonid Levin. This subject is frequently taught in computer science courses. Beyond just computer scientists, Chaitin’s research has garnered the interest of numerous philosophers and mathematicians, leading them to explore profound questions in mathematical innovation and digital philosophy.

Gregory Chaitin is author of Information, Randomness & Incompleteness (1987) Algorithmic Information Theory (1987) The Limits of Mathematics (1998), The Unknowable (1999), Exploring Randomness (2001), Meta Math!: The Quest for Omega (2005), Mathematics, Complexity and Philosophy (2011), Gödel’s Way (2012) and Proving Darwin: Making Biology Mathematical (2012).

_________________________________

Gregory Chaitin a exploré profondément les métamathématiques, les considérant à la fois comme alliées et rivales des mathématiques traditionnelles. Les métamathématiques, qui scrutent les mathématiques à travers un prisme mathématique, trouvent leurs racines chez Leibniz, leur précurseur, et David Hilbert, leur principal représentant. Elles s’appuient principalement sur des systèmes axiomatiques formels, qui sont des représentations élevées et structurées des mathématiques. Ces systèmes sont les fondations du raisonnement mathématique, strictement encadrés pour éliminer toute subjectivité grâce à la logique symbolique.

Chaitin a tracé l’histoire des métamathématiques, citant Hilbert qui voyait les mathématiques comme étant fondamentalement dichotomiques —  soit vraies, soit fausses —  lorsqu’elles sont pratiquées dans un cadre axiomatique formel. Cela nécessite une structuration et une formalisation précises, comme le montrent les contributions d’Alfred  Tarski en géométrie. Le système arithmétique de Peano et la théorie des ensembles de Zermelo-Fraenkel sont des exemples notables de tels cadres.

Chaitin a présenté l’idée du “programme élégant”, un code épuré générant des résultats déterminés, le reliant au théorème d’incomplétude. Il a mis en évidence les défis de prouver un tel programme, montrant les limites des systèmes axiomatiques formels pour cette mission. L’incomplétude, qui remet en question l’idée d’un système mathématique universel, reste un thème récurrent dans son argumentation.

Chaitin a conclu en affirmant que tout système axiomatique, aussi élaboré soit-il, ne peut appréhender qu’une minuscule partie des immenses complexités des mathématiques pures platoniciennes. Il a insisté sur la nécessité de repenser les axiomes et d’embrasser de nouveaux paradigmes pour aborder le mystère de l’indécidabilité et stimuler l’avancement mathématique. Bien que ce sujet soit dense et empreint de réflexions philosophiques, il offre une vision élargie de l’évolution des mathématiques et de leur relation à la pensée humaine.

La discussion post-présentation poursuivie avec Stephen Wolfram, Hervé Zwirn et David Chavalarias a permis de fusionner les opinions de différents experts, oscillant entre espoir et réserve sur le futur des mathématiques. Les réflexions ont varié de la synergie entre mathématiques et science malgré l’incomplétude, à l’importance de l’incompressibilité informatique dans l’essence de la réalité. L’intervention de Chaitin a brillamment souligné le lien entre les mathématiques, la pensée et l’infini voyage de la découverte humaine.

H.A

Gregory Chaitin, né en 1947, est un mathématicien et informaticien argentin-américain. Depuis la fin des années 1960, il a apporté d’importantes contributions à la théorie de l’information algorithmique et à la métamathématique, notamment un résultat notable en théorie informatique analogue au théorème d’incomplétude de Gödel. Chaitin, avec Andrei Kolmogorov et Ray Solomonoff, est considéré comme l’un des pionniers du domaine de la complexité algorithmique, souvent appelée complexité de Solomonoff-Kolmogorov-Chaitin, ou complexité de taille de programme. L’émergence de la théorie de l’information algorithmique comme aspect central de l’informatique théorique, de la théorie de l’information et de la logique mathématique doit beaucoup aux contributions de figures comme Solomonoff, Kolmogorov, Martin-Löf et Leonid Levin. Au-delà du domaine informatique, les recherches de Gregory Chaitin ont suscité l’intérêt de nombreux philosophes et mathématiciens, les incitant à explorer des questions profondes d’innovation mathématique et de philosophie digitale.

Gregory Chaitin est l’auteur de Information, Randomness & Incompleteness (1987) Algorithmic Information Theory (1987) The Limits of Mathematics (1998), The Unknowable (1999), Exploring Randomness (2001), Meta Math!: The Quest for Omega (2005), Mathematics, Complexity and Philosophy (2011), Gödel’s Way (2012) et de Proving Darwin: Making Biology Mathematical (2012).

_________________________________

Sara Zouiten, ‘World-Acclaimed Scientists Tackle ‘Complexity’ at UM6P Science Week’, Morocco News, Feb. 20, 2023
https://www.moroccoworldnews.com/2023/02/354130/world-acclaimed-scientists-tackle-complexity-at-um6p-science-week

Rabat – The third edition of the UM6P Science Week kicked off on Monday at Mohammed VI Polytechnic University’s campus in Ben Guerir, bringing together high-caliber scientists, thinkers, and researchers from across the world.

The event, which is taking place between February 20 and 26, features a set of conferences, musical performances, reading circles and other activities, with each participant tackling the theme of “complexity” in relation to their field of expertise.

The UM6P Science Week “plays a catalytic role in the life of the university,” Moroccan professor and sociologist Reda Benkirane emphasized in an interview with Morocco World News, adding that the event offers students and participants the opportunity to exchange in a “multidisciplinary manner” with “the greatest scientists of our time.”

The ‘beginning of something beautiful’

Speaking at the opening of the event, UM6P President Hicham El Habti also highlighted the importance of the university’s science week, saying it aligns with the UM6P’s vision to promote applied scientific research.

El Habti hopes that the event would be “the beginning of something beautiful,” and contribute to creating a “more realistic impact.” In this regard, he called on the participants and UM6P students to seize the opportunity and benefit from the expertise and knowledge of the participating scientists. “Dare to explore,” he stressed.

Explaining the reasoning behind this edition’s theme, Moroccan academic and writer Fouad Laroui emphasized that complexity is “common in all domains” and is present in the entirety of the human experience.

Moroccan academic Fouad Laroui speaking at the event

Benkirane, who is also the Chair of Complexities and Humanities at the Africa Business School, echoed Laroui’s statement, emphasizing the relevance and presence of complexity in all components of humanity. “Complexity is all around us,” he said.

However, Benkirane warned that the complexity in play could have a reverse impact in the future, in a “butterfly effect” fashion. “Complexity is somehow taking revenge on us,” he said, highlighting the many crises and challenges that humanity has faced over the past decades, from 9/11 to the covid-19 pandemic. “We are living in a dark age of complexity.”

Under the theme of “Complexity, Golden Age, Dark Age,” Benkirane’s presentation indicated that regardless of its future impact, which could be better or worse, complexity will undoubtedly be more “amplified and generalized .”

Chair of Complexities and Humanities at the Africa Business School Reda Benkirane

Complexity in Times of Crisis

Like Benkirane, French philosopher and sociologist Edgar Morin, spoke about the “challenges of complexity in times of crisis.” The well-respected 101-year-old social scientist has witnessed some of humanity’s biggest crises, challenges, and prominent events, including his participation in World War II.

“One cannot understand a system, unless one understands the properties of all its components,” Morin emphasized, explaining that life does not follow a linear order. The French sociologist also highlighted the complexity of the relations that govern people and institutions, saying that “any autonomy is also dependency.”

Several unexpected scenarios unfold in times of uncertainty and interlinked crises, Mourin noted, explaining that difficult times can prompt creativity and innovative solutions, as well as have severe repercussions and trigger other crises.

French philosopher and sociologist Edgar Morin

Talking about the crises that have engulfed the world in recent decades, Mourin said that they tend to pile up and come one after the other. He gave the example of the COVID-19 pandemic and the war in Ukraine, and the effect they have had on globalization. Describing it as “the crisis of globalization,” he said that the ongoing war has created obstacles in the face of the years-long efforts to unite the world.

In addition, the complexity of the COVID-19 crisis provoked several social, domestic, and economic crises, such as the gas and energy crisis and food insecurity in Africa. “The crisis happened during a time when the planet was in crisis,” he said in reference to the challenges that currently face humanity, such as climate change, which has aggravated in light of the current global context.

Morin drew attention to the difficult dilemma humans are bound to encounter as they grapple with these successive crises that are happening simultaneously and are affecting several domains. “This is humanity in 2023 … encountering crises that are inflating one another.”

Read also: A Panel’s Fun Foray Into the Tragic Geopolitics of Climate Change 

“Humanity has no way of managing them [the challenges],” the French scientist said, stressing that “no entity can manage this crisis.” He argued that “war creates illusion and hatred,” encouraging people to promote and engage in a globalization that is based on “solidarity.”

Since all humans are affected by these crises, they must come together in pursuit of collective action, he stressed.

Meanwhile, Stephen Wolfram, a British-American scientist and CEO of Wolfram Research, spoke of the “foundation and implications of complexity.” He discussed his concept of the “Ruliad,” which is “the result of following all possible computational rules in all possible ways.”

British-American scientist Stephen Wolfram

Explaining the complexity of the universe, Wolfram said that “the behavior could be simple, but the result we get from it is complex.” “The universe is running by all possible rules,” he added, explaining that it is difficult to pinpoint the very structure that governs the world.

Wolfram gave the example of ChatGPT – an Artificial Intelligence powered chatbot – during his presentation, saying that the AI-powered tool is capable of generating human-like language by implicitly learning the syntactic grammar of English.

While many think that a complex process comes into play to produce ChatGPT’s human-like language, the English-American scientist said that it simply uses the humans’ repeated structure of meaning and generalization of logic.

Other participants at the event include French philosopher, sociologist, and professor Raphael Liogier, who told Morocco World News about the goal he aims to achieve through his participation in the UM6P Science Week. He said he seeks to fight against two widely spread prejudices.

French professor and sociologist Raphael Liogier

One of those prejudices is the belief that human beings’ intelligence and conscience are the result of their complexity, Liogier explained.

The other prejudice Liogier aims to debunk is humans’ belief that they are more complex than other animals. For him, humans are in fact “complexed” animals that seek to compensate for their fragility and weakness. He told MWN: “We are complexed before we are complex.” 

Meanwhile, President of the Makassed University of Beirut Hassan Ghaziri will tackle the way the “notion of complexity is transforming the engineering areas,” as well as its role in finding new ways to solve our current issues, he told Morocco World News.

Ghaziri, who is also a professor of machine learning, argued that engineers are the “agent of transformation” in our current world, explaining that at this moment the world faces “deep crises” due to new technologies, including Artificial Intelligence (AI).

Professor of machine learning and President of the Makassed University of Beirut Hassan Ghaziri

© Morocco World News. All Rights Reserved.

__________________________________

لقاء مفكري التعقيد

ما الذي يشترك فيه تعقيد شبكة الحاسوب وتعقيد الحشرات الاجتماعية؟ ما هي إمكانيات وحدود الذكاء الاصطناعي؟ هل يمكننا إعادة اختراع الحياة، والتحكم في التقاطعات الجينية؟ ما الوعود والمخاطر التي يجب أن نراها في ارتفاع علم الأحياء؟ هل الغموض وعدم القدرة على التنبؤ والفرص هي مظاهر لجهلنا أم جوانب أساسية؟ هل الله يلعب بالنرد؟ هل الكون محدود أم لا نهائي، مؤقت أم أبدي؟ هل يمكن أن يكون هناك “نظرية لكل شيء”؟

هذه بعض الأسئلة المثيرة التي طرحتها هذه القصص العلمية. 

يقدم الكتاب أيضًا معرضًا لصور العلماء من خلاله يصبح جودة الشهادات و، قبل كل شيء، تنوع الآراء المجمعة، واضحة. خلف بعض أعظم أسماء العلم، تظهر الأفراد، مكشفة عن حساسياتهم، الاختلافات، الخلافات، الحدود، والآمال. في النهاية، يروي كل منهم قصتهم بهذه الطريقة، تحت نظرة رضا بنكيران ، التي تكون أحيانًا متواطئة وأحيانًا نقدية.

تم تقديم هذه الطبعة بمقدمة من الكاتب تحدد الكتاب فيما يتعلق بالحالة الحالية للبحث عن التعقيد. يتكون الكتاب من مقدمة، وثمانية عشر فصلاً، وقائمة مراجع مذيلة.

مقابلات مع، حسب ترتيب الظهور: إدغار موران، إيليا بريغوجين، نيل جيرشنفيلد، دانيال مانج، جان-لوي دينوبورغ، لوك ستيلز، كريستوفر لانغتون، فرانسيسكو فاريلا، براين جودوين، ستيوارت كوفمان، برنارد ديريدا، إيف بوميو، إيفار إيكيلاند، غريغوري شايتين، جون بارو، لوران نوتال، أندري ليند، ميشيل سيريس.

__________________________________

سوسيولوجي، حاصل على دكتوراه في الفلسفة، رضا بنكيران خبير في العلاقات الدولية في الأمم المتحدة والمنظمات غير الحكومية في جنيف. هو أستاذ مرتبط بالكامل في كلية الأعمال الأفريقية لجامعة محمد السادس التقنية، وهو مسؤول عن كرسي التعقيد ∞ الإنسانيات. يدرس أيضًا في جامعة فريبورغ في سويسرا. ركزت أبحاثه التجريبية بشكل رئيسي على الحضرية، التعقيد، التطرف، والإسلامية.