On 25 March 2026, Émile Servan-Schreiber reshared a photograph that Rhoda Fry said she had taken at Carnegie Mellon University in 1983. It showed Steve Jobs and Jean-Jacques Servan-Schreiber listening to a student’s question about personal computers. Émile recalled being there. What stayed with him, he wrote, was the concentration of the two men and the experience of being around original thinkers.
He linked that memory to a lifetime studying and building intelligent systems: cognitive, collective and artificial. This was a personal account of inspiration. Two earlier posts give it more substance, tracing his interest in collective behaviour through an undergraduate AI project and a graduate-school choice.
An AI project about societies of agents
In a post dated 2 May 2023, Servan-Schreiber recalled taking Geoffrey Hinton’s AI class at Carnegie Mellon in 1984. His term project concerned emergent collective behaviour in societies of agents: how interactions among separate actors can produce behaviour at the group level. He said Hinton awarded it a B+ and suggested using more AI techniques. His recollection places the question of group behaviour near the beginning of his studies, before his later work with forecasting crowds.
Raj Reddy’s unexpected recommendation
A reunion in Paris prompted a more consequential recollection on 11 November 2021. Servan-Schreiber said that in 1986 he had asked Raj Reddy whether to pursue artificial intelligence or robotics in graduate school. Reddy recommended a PhD in cognitive psychology instead, with John R. Anderson and Herbert Simon; Servan-Schreiber also named Allen Newell as his third adviser.
Carnegie Mellon’s faculty record confirms that Reddy was the Robotics Institute’s founding director from 1979 to 1991, covering the period of that conversation. Anderson’s university profile describes his work on ACT-R, a computational architecture intended to model human cognition. Reddy’s recommendation and its effect on his career remain Servan-Schreiber’s recollections.
From individual minds to collective judgement
He wrote that he had never considered psychology before Reddy’s recommendation, then quickly fell in love with cognitive science. In the 2021 account, he connected that interest directly to Hypermind and to his teaching and research at the UM6P School of Collective Intelligence. The 2026 post widened the circle of influences to Hinton and James McClelland as well as Simon and Anderson.
The connection matters because forecasting asks both kinds of question: how individuals form judgements, and how a system combines those judgements. His later writing on collective intelligence explores that relationship. His intellectual route connects the study of individual minds with a practical problem for forecasting: designing a process in which different people’s information can contribute to a common judgement.
