Shane Legg (born in New Zealand) is a renowned machine learning scientist and researcher, co-founder of DeepMind Technologies and current Chief AGI Scientist at Google DeepMind.
He is widely recognized for having formulated one of the first rigorous mathematical definitions of Artificial General Intelligence (AGI), guiding the scientific goals and security protocols that govern the industry today.
Mathematical foundations of machine intelligence
Legg earned his PhD at the renowned Dalle Molle Institute for Artificial Intelligence Studies (IDSIA) in Lugano, Switzerland, under the direction of Professor Marcus Hutter. His doctoral thesis, awarded the Canadian Singularity Institute Research Prize, theoretically formalized general intelligence in autonomous agents using concepts of Kolmogorov complexity and algorithmic information theory.
The founding of DeepMind and its arrival at Google
In 2010, Legg joined forces with Demis Hassabis and Mustafa Suleyman to found DeepMind Technologies in London, with the explicit ambition of solving artificial intelligence and then solving everything else. In 2014, DeepMind was acquired by Google, becoming the epicenter of advanced AI research.
Since the birth of DeepMind, Legg has been the leading advocate for technical safety in AI. He established the company’s AGI Safety group and has driven standards to ensure that models with superintelligence capabilities remain aligned with core human values and well-being.
AGI Safety leadership and ethical alignment
His work has decisively influenced the development of milestones such as AlphaGo, AlphaZero and the recent Gemini models, positioning him as one of the most distinguished minds on the path to artificial general intelligence.
Explore more profiles in our directory: Meet all the scientists, founders and leaders who are setting the course for technology in our pillar guide: Who’s who in AI?.
Cet article est également disponible en :