Education
-
École supérieure d’ingénieurs en électrotechnique et électronique (ESIEE Paris):
He received his Diplôme d’Ingénieur (equivalent to a Master’s in Engineering) in 1983. -
Pierre and Marie Curie University (now part of Sorbonne University), Paris, France:
He earned his PhD in Computer Science in 1987.
His doctoral thesis was on “Modeles connexionnistes de l’apprentissage” (Connectionist Models of Learning), supervised by Maurice Milgram.
Activities
historical
- Pioneering Deep Learning: Developed convolutional neural networks (CNNs) in the late 1980s and 1990s, foundational for modern computer vision.
- Handwriting Recognition: Created systems for handwriting recognition, notably used by banks for reading checks.
- Academic Roles: Professor at New York University (NYU), contributed to machine learning research and education.
- AT&T Bell Labs: Worked on neural networks and machine learning during the 1990s.
present
- Chief AI Scientist at Meta (Facebook): Leading AI research, focusing on self-supervised learning, energy-based models, and advancing AI capabilities.
- NYU Professor: Continues to teach and supervise research at NYU’s Center for Data Science.
- Public Speaker & Advocate: Frequently speaks at conferences and in media about the future of AI, ethics, and research directions.
Connections to other people and companies
- Meta (Facebook): Chief AI Scientist.
- NYU: Professor of Computer Science, Data Science, Neural Science.
- Geoffrey Hinton & Yoshua Bengio: Collaborators and co-recipients of the 2018 Turing Award; together known as the “Godfathers of Deep Learning.”
- AT&T Bell Labs: Former employer.
- Element AI: Advisory connections (company co-founded by Bengio).
- AI Research Community: Active collaborator with leading researchers worldwide.
Expectations for the future
- Self-Supervised Learning: Predicts this will be key to achieving more general and autonomous AI systems.
- AI Safety & Ethics: Advocates for responsible development but is skeptical of near-term existential risks from AI.
- General Intelligence: Believes progress toward human-level AI will be gradual and driven by advances in learning algorithms rather than scaling up current models alone.
- Open Science: Supports open publication and sharing of research to accelerate progress.
Interests
- Robotics: Interest in applying deep learning to robotics and embodied intelligence.
- Music & Art: Occasionally discusses intersections between AI, creativity, music, and art.
- Education & Outreach: Passionate about teaching, mentoring, and making AI accessible to broader audiences.
- Science Communication: Active on social media platforms, engaging with both technical and general audiences about AI topics.
