The artificial intelligence tools we use today, including Siri, Google Translate or ChatGPT, would not exist if pioneers such as the computer scientist Yorick Wilks had not helped to establish the field of natural language processing: teaching computers to interpret, generate and translate human language. Crucial to Wilks’s research and career progression was his experience in Stanford, California, where he worked in the lab of the AI pioneer John McCarthy in the 1970s.
Wilks, who has died aged 83, undertook work in computational linguistics, machine translation and AI more broadly. One of his achievements was the development of the preference semantics model, a technique for representing the meaning of words and phrases by considering their context and usage in natural language texts. This approach has found extensive use in automated question-answering systems such as chatbots.
In 1997 Wilks served as the chief researcher of the group led by the British chess player and computer expert David Levy that won the Loebner prize for machine dialogue, awarded to the most human-like conversational computer program. Their chatbot, Catherine, was designed to mimic the conversational style of an English journalist. Wilks recalled: “We made her British because, if she made any mistakes in New York [where the competition was judged that year], they might think it’s because she was British.”
In his later research, Wilks delved into the concept of artificial companions: conversational agents designed to interact with elderly people or other isolated individuals using speech, learning their tastes and habits, or reminding them of their medications. He later imagined that chatbots and other digital companions could use AI to mimic the voice and learn the memories of people in order to impersonate them. This could even enable relatives to interact with their loved ones after their death. Conscious of the ethical implications of AI, Wilks discussed the issue in a series of public lectures in 2018-20, when he was visiting professor of AI at Gresham College, London.
Yorick was born in Gerrards Cross, Buckinghamshire, where his mother, Peggy (nee Weinel), was staying at the time, a few weeks after the second world war began, but he grew up in Edmonton, north London. Peggy worked as a hotelier, chef and aircraft inspector, and his father, Alexander Wilks, was a carpenter and joiner. Yorick was 11 when his father died. The family then moved to Devon; Yorick was educated at Torquay boys’ grammar school, and won a scholarship to study physics at Pembroke College, Cambridge, in 1958.
He later changed his programme of study, first to mathematics and then to philosophy, entering the circle of Margaret Masterman’s epiphany philosophers, who focused on the relationship between science and religion, and working in the Cambridge Language Research Unit on early programs to do syntax analysis and text extraction.
During his Cambridge years Wilks developed a talent for theatre and a passion for politics. Later in life he continued to perform in amateur theatre and to be an active commentator on politics and public affairs, sparing no wing of any party from criticism. He became a member of the Reform Club in central London in 2007, and served as an adviser on AI-related issues to the Centre for Policy Studies.
In 1966 he left Cambridge for Los Angeles, thanks to a job which enabled him to work on more advanced computers. After the end of his contract, he stayed in California, supporting himself by playing a small part as a comedian in a TV show, while writing his PhD dissertation and getting his doctorate from Cambridge in 1968. The following year he became a research associate in the Stanford Artificial Intelligence Laboratory, where he worked on machine translation programs.
In 1974 he moved back to Europe, joining the Dalle Molle Institute for Semantic and Cognitive Studies in Lugano, Switzerland, a centre for the application of AI to linguistics and automated translation. The focus of his research then shifted to belief systems: how humans need a model of the beliefs of another person in order to communicate with them.
After a short period at the University of Edinburgh, in 1976 he moved to the University of Essex where he eventually became professor of linguistics and computer science, working on the large-scale Eurotra machine translation project. Wilks spoke French, German, Italian, Spanish, Swahili and Japanese.
In 1985 he moved back to the US to head the computing research lab at New Mexico State University, Las Cruces, where he worked on the development of a state-funded AI laboratory, doing early work on information extraction systems. In 1998 he became head of the department of computer science at the University of Sheffield, where he had started working in 1993 as professor of AI.
Wilks continued his professional relations with the US after moving to Oxford in 2003 and leading the large EU-funded Companions project at the Oxford Internet Institute; at the age of 70 he joined the Florida Institute for Human and Machine Cognition, where he established a new AI group to research cybersecurity, and belief and emotion propagation in groups – how, for instance, changes in ways of thinking can be detected in the use of language on social media platforms. He had recently completed a final book, Artificial Intelligence and God, to be published by Oxford University Press.
Wilks’s work was recognised by awards including the lifetime achievement award of the Association for Computational Linguistics, and, in 2009, the Lovelace medal of the British Computer Society.
Wilks is survived by his third wife, Roberta Catizone, a fellow researcher in AI, whom he married in 1993, and their children, Octavia and Zoe; by two children, Seth and Claire, from his second marriage, to Geraldine de Berly, which ended in divorce; by two grandchildren; and by his brother, Leif. His first wife, Felicity Ann Snee, a doctor, died in the 1970s.
• Yorick Alexander Wilks, computer scientist, born 27 October 1939; died 14 April 2023