(Photo: provided.)
From co-authoring seminal research papers to co-founding the research organization that developed ChatGPT, few people have been as influential in shaping the artificial intelligence landscape – and conversations around the technology’s responsible use — as Ilya Sutskever.
As a University of Toronto graduate student, Sutskever co-authored one of the most cited academic papers of this century and has since played a central role in driving the development and adoption of a technology that is transforming the economy, society and people’s everyday lives.
Today, for his foundational work and global impact as a computer scientist and artificial intelligence (AI) visionary, and for his outstanding service as an advocate of safe and responsible AI, Sutskever will receive a Doctor of Science, honoris causa, from U of T.
Born in Russia and raised in Israel, Sutskever became fascinated with computing at age five, when he first laid eyes on a computer — “I was utterly enchanted,” he told U of T Magazine in 2022 — and his interest continued into his teen years, when he emigrated to Canada with his family.
Even as a teenager, Sutskever envisioned building computers with human-like capabilities. “I remember thinking a lot about the nature of existence and consciousness … about souls and intelligence. I felt very strongly that learning was this mysterious thing: humans clearly learn, computers clearly don’t.”
Admitted into U of T’s math program out of Grade 11, Sutskever immediately immersed himself in upper-year courses. Graduating with an honours bachelor of science degree in mathematics in 2005, he went on to earn a master’s degree and PhD in computer science at U of T — the latter under the supervision of University Professor Emeritus Geoffrey Hinton, winner of the 2024 Nobel Prize in Physics.
Hinton recalls being hugely impressed with Sutskever in their early interactions. In an interview with the Globe and Mail, the “godfather of AI” recounted giving Sutskever — who had knocked on his door and expressed an interest in joining his lab — a paper to read and being taken aback by the clarity of his responses. “His immediate reaction to things were reactions that had taken experts in the field quite a long time to come up with,” Hinton said.
Among Sutskever’s research projects at U of T was a program that used neural networks, which are computational models inspired by the human brain, to learn about language and generate text — a crude forerunner to ChatGPT. “I give it an initial segment of text. And I say, from this text, keep on producing text that you think looks like Wikipedia,” he told U of T Magazine in 2010.
Then, in 2012, Sutskever, Hinton and another of Hinton’s graduate students, Alex Krizhevsky, developed AlexNet, a convolutional neural network that was trained to identify objects in a purpose-built image database with far more accuracy than competing approaches — effectively changing the AI game overnight. (The source code for AlexNet is to be preserved at the Computer History Museum in Silicon Valley.)
Sutskever then joined Hinton’s spinoff company DNNResearch, which was later acquired by Google. Brought on as a research scientist at Google Brain, he contributed to yet another AI milestone: Training a computer program called AlphaGo, powered by deep neural networks, to play the ancient strategy game of GO — and then beating a professional (human) player. He also co-developed sequence-to-sequence models, which are foundational to current machine translation systems.
In 2015, Sutskever co-founded OpenAI, serving as its research director and later as chief scientist. Under his leadership, OpenAI introduced the large language models that power ChatGPT, the generative AI chatbot now used by millions around the world for everything from drafting emails and sourcing recipes to writing computer code. And he played a central role in the creation of large reasoning models, which perform complex reasoning tasks.
Sutskever left the organization last year and co-founded Safe SuperIntelligence, a company that is developing safe AI systems with superhuman capabilities.
“We plan to advance capabilities as fast as possible while making sure our safety always remains ahead,” Sutskever and co-founders said in a statement announcing the venture.
Sutskever’s achievements have led to him being elected to the prestigious Royal Society in the UK and being named among TIME’s 100 Most Influential People in AI in 2023 and 2024, among other honours.
— Original story by Rahul Kalvapalle for U of T News