A summary of the presentation made at the 2024 Eye Data and AI Summit
By Professor Jimmy Lin
Professor, and David R. Cheriton Chair in Software Systems
Cheriton School of Computer Science, University of Waterloo
Co-director, Waterloo Data & Artificial Intelligence Institute, Primal
Understanding ChatGPT and Large Language Models (LLMs)
Artificial intelligence (AI) tools like ChatGPT are changing how we interact with technology. These tools can understand and respond in human-like language, making it easier to communicate, learn and create.
What Can ChatGPT Do?
ChatGPT is built on a type of artificial intelligence called a large language model (LLM). It istrained in vast amounts of text and can generate responses that are helpful, creative, and even impressive. People use it from writing, summarizing and explaining complex information. Its ability to mimic human conversation has made it one of the most talked-about technologies in recent years.
Challenges and Limitations
Despite its strengths, ChatGPT is not perfect. It can sometimes produce incorrect or misleading information—known as “hallucinations”—or reflect biased views based on the data it was trained on. These issues raise concerns about fairness, accuracy, and trust.
Risks of Misuse
When AI tools like ChatGPT are used without proper oversight, they can contribute to the spread of false information or reinforce harmful stereotypes. There are also concerns about job displacement and the ethical use of content created by others. Without clear guidelines and accountability, these risks can grow.
Opportunities with Responsible Use
When used thoughtfully and within a strong governance framework, ChatGPT and similar tools can offer real benefits everywhere. For example, in healthcare, they can support healthcare professionals by streamlining access to information, improving workflows and enhancing patient experience. It provides better coordination and decision-making. The key is to ensure these tools are aligned with human values and used to support—not replace—human judgment.
Looking Ahead
AI technologies like ChatGPT are here to stay, and their impact will continue to grow. By understanding both their potential and their limitations, we can make informed choices about how to use them responsibly. With the right safeguards in place, these tools can help us solve problems, improve services, and create new opportunities across society.