Yorick Wilks, a pioneer in natural language processing and artificial intelligence, passed away at the age of 83. Wilks was instrumental in teaching computers to interpret, generate, and translate human language, and his contributions were crucial in the development of modern AI tools such as Siri, Google Translate, and chatbots like ChatGPT. His career was shaped by his experience working in the lab of AI pioneer John McCarthy at Stanford University in the 1970s.
Wilks’ research focused on computational linguistics, machine translation, and AI more broadly. He developed 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 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.
In his later research, Wilks explored 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 even 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.
Wilks was born in Gerrards Cross, Buckinghamshire, in 1939, where his mother, Peggy (nee Weinel), was staying at the time, a few weeks after the second world war began. He grew up in Edmonton, north London, where his mother 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, and the family then moved to Devon. He 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, where he worked 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 center 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. He had recently completed a final book, Artificial Intelligence and God, to be published by Oxford University Press.
Wilks was recognized for his work with awards including the lifetime achievement award of the Association for Computational Linguistics and the Lovelace medal of the British Computer Society in 2009. He 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.