Yoav Shoham
Last reviewed
May 31, 2026
Sources
16 citations
Review status
Source-backed
Revision
v1 · 1,835 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
May 31, 2026
Sources
16 citations
Review status
Source-backed
Revision
v1 · 1,835 words
Add missing citations, update stale details, or suggest a clearer explanation.
Yoav Shoham is an Israeli and American computer scientist known for foundational research in multi-agent systems and the use of game theory and formal logic in artificial intelligence, and for a career in technology entrepreneurship. He is professor emeritus of computer science at Stanford University, where he taught for about 28 years. In 2017 he co-founded AI21 Labs with Amnon Shashua and Ori Goshen, a company that builds large language models and language tools, including the Jurassic series, the Jamba model, and the Wordtune writing assistant. Shoham also conceived the AI Index, an annual report that tracks progress in the field, and he is a Fellow of the Association for Computing Machinery, the Association for the Advancement of Artificial Intelligence, and the Game Theory Society. [1][2][3]
Before AI21 Labs he founded or co-founded several earlier startups, among them TradingDynamics, Katango, and Timeful. The latter two were acquired by Google, where he then worked as a principal scientist. With Kevin Leyton-Brown he wrote the widely used graduate textbook "Multiagent Systems: Algorithmic, Game-Theoretic, and Logical Foundations." [1][4]
Yoav Shoham was born in Israel on January 22, 1956. [1] He studied at the Technion, the Israel Institute of Technology, where he earned a bachelor of science degree. [1][3]
He then moved to the United States for doctoral study at Yale University. His doctoral work was in computer science and focused on formal methods for reasoning about time, change, and causation, an area at the boundary of logic and artificial intelligence. His adviser was Drew McDermott, a researcher in knowledge representation and temporal reasoning. [1][5] Shoham received his Ph.D. from Yale in 1987, and the published version of his dissertation appeared in 1988 as the book "Reasoning about Change: Time and Causation from the Standpoint of Artificial Intelligence," issued by MIT Press. [1][5]
Shoham joined the computer science faculty at Stanford University, where he spent roughly 28 years and rose through the academic ranks before taking emeritus status. His official Stanford title is professor of computer science, emeritus. [2][3] At Stanford he worked within the artificial intelligence group and the robotics laboratory, and he supervised doctoral students who went on to research careers of their own, including Kevin Leyton-Brown, his later co-author. [1][2]
His research over this period ranged across several connected areas: formal logics of knowledge and belief, nonmonotonic and temporal reasoning, the design of software agents, and the application of game theory and economic mechanism design to computational settings. [2][3] The thread running through much of this work is the question of how to give a precise mathematical account of agents that hold beliefs, make decisions, and interact with one another.
Shoham's early research continued the line of his dissertation. He studied how a computer system can represent and reason about events that unfold over time, including which conclusions should be retracted when new information arrives, a problem that belongs to the study of nonmonotonic reasoning. He also contributed to the formal logic of knowledge and belief, sometimes called epistemic logic, which models what an agent knows, what it knows about what other agents know, and how that knowledge changes. This formal machinery later became part of the toolkit he applied to systems of many interacting agents. [2][5]
In the early 1990s Shoham proposed a programming framework he named agent-oriented programming. He set it out in the paper "Agent-Oriented Programming," published in the journal Artificial Intelligence in 1993. [6] The idea treats a program as a society of agents whose internal state is described in mental terms such as beliefs, decisions, capabilities, and obligations, with these notions given formal definitions in an extended logic of knowledge and time. Agents in the framework communicate through structured messages drawn from speech act theory, with message types such as informing, requesting, and offering. [6] The proposal influenced later work on agent programming languages and helped frame the broader research program on software agents.
A large part of Shoham's later academic work brought game theory and microeconomic ideas into computer science. He studied how self-interested agents can reach agreements, how to design auctions and other mechanisms whose rules lead to good outcomes even when participants act strategically, and how groups of agents can coordinate or learn. Topics in this body of work include combinatorial auctions, social choice, cooperative game theory, and multiagent learning. [2][3] This research sits within the field now often called algorithmic game theory, which studies the computational side of strategic interaction.
With Kevin Leyton-Brown of the University of British Columbia, Shoham wrote "Multiagent Systems: Algorithmic, Game-Theoretic, and Logical Foundations," published by Cambridge University Press in 2009. [4][7] The book gathers into one volume the mathematical foundations of systems in which many agents act and interact, covering distributed problem solving, game theory, auctions and mechanism design, social choice, multiagent learning, and logics of knowledge and belief. It became a standard graduate reference for the field, and the authors made the full text available free online. [7][8]
The two authors also produced a shorter companion text, "Essentials of Game Theory: A Concise, Multidisciplinary Introduction," published in 2008. [1] Alongside the books, Shoham co-taught open online courses on game theory through Coursera with Leyton-Brown and the economist Matthew O. Jackson of Stanford. The courses reached a very large audience, with Shoham's profile describing his online game theory teaching as watched by close to a million people. [2][3]
Shoham built companies in parallel with his academic work, repeatedly turning research ideas into products. [1][3]
In 1999 he founded TradingDynamics, a business-to-business commerce company that developed software for online trading and marketplaces. The company was acquired by Ariba in 2000. [1] In 2011 he co-founded Katango, which used algorithms to help people organize their contacts and social circles; Google acquired it in 2013. [1] In 2014 he co-founded Timeful, a time-management and scheduling startup whose collaborators included the behavioral economist Dan Ariely. Google acquired Timeful in 2015, and Shoham then joined Google as a principal scientist, a role he held until 2017. [1][9]
In 2017 Shoham co-founded AI21 Labs with Amnon Shashua, the computer scientist and entrepreneur behind Mobileye, and Ori Goshen. The company is based in Tel Aviv, Israel, and develops language models and natural language tools. [1][10] Shoham has served as co-chief executive of the company. [10]
AI21 Labs released the Wordtune writing assistant in 2020, a consumer product that suggests rewrites and paraphrases and that can summarize text. [10] In 2021 the company introduced Jurassic-1, one of the larger language models available at the time, and it followed with Jurassic-2 in 2023. [1][10] In March 2024 AI21 Labs released Jamba, an open-weights model built on a hybrid architecture that combines the Mamba structured state space model with elements of the Transformer design and a mixture-of-experts layer. Jamba supported context windows of up to 256,000 tokens and was released under the Apache 2.0 license, and the company later released updated Jamba 1.5 versions. [11][12] The company raised venture funding across several rounds, including a Series C in 2023 with investors that included Google and Nvidia, which valued the company in the range of well over one billion dollars. [10][13]
Shoham conceived the AI Index, an effort to measure and report on activity and progress in artificial intelligence using data rather than opinion. He assembled a founding steering committee whose members included Ray Perrault of SRI International, Erik Brynjolfsson of the Massachusetts Institute of Technology, and Jack Clark, then of OpenAI. The project grew out of the One Hundred Year Study on Artificial Intelligence and was launched publicly at the end of 2017. [1][14] At its start Shoham described the goal as providing a fact-based measuring stick against which to chart progress and inform discussion about the future of the field. [14] The AI Index later became a program of the Stanford Institute for Human-Centered Artificial Intelligence, which now publishes its annual report. [14][15]
Shoham has also chaired WeCode, a nonprofit initiative that trains programmers from disadvantaged backgrounds. [2][3]
Shoham is a Fellow of three scholarly bodies: the Association for Computing Machinery, the Association for the Advancement of Artificial Intelligence, and the Game Theory Society. [2][3] His individual awards recognize contributions to multiagent systems and artificial intelligence. [2][16]
In 2008 he received the ACM/SIGART Autonomous Agents Research Award. In 2013 he and a co-recipient were given the ACM-AAAI Allen Newell Award, an honor for research that bridges computer science with other disciplines. In 2019 he received the IJCAI Award for Research Excellence, presented at the International Joint Conference on Artificial Intelligence. [2][3][16]
| Field | Detail |
|---|---|
| Full name | Yoav Shoham |
| Born | January 22, 1956, Israel |
| Nationality | Israeli, American |
| Fields | Artificial intelligence, multi-agent systems, game theory, logic |
| Undergraduate | Technion, Israel Institute of Technology (B.Sc.) |
| Doctorate | Yale University, Ph.D. in computer science, 1987 |
| Doctoral adviser | Drew McDermott |
| Institution | Stanford University (professor of computer science, emeritus) |
| Doctoral student | Kevin Leyton-Brown |
| Known for | Agent-oriented programming, multiagent systems textbook, co-founding AI21 Labs, the AI Index |
| Companies founded | TradingDynamics, Katango, Timeful, AI21 Labs |
| Major textbook | "Multiagent Systems: Algorithmic, Game-Theoretic, and Logical Foundations" (2009, with Kevin Leyton-Brown) |
| Fellowships | ACM Fellow, AAAI Fellow, Game Theory Society Fellow |
| Selected awards | ACM/SIGART Autonomous Agents Research Award (2008), ACM-AAAI Allen Newell Award (2013), IJCAI Award for Research Excellence (2019) |