Professor James Mclaurin is Co-director for the Centre of AI and Public Policy at the University of Otago. He spoke to members at our meeting on Thursday 16 Nov. Here is a detailed description of his talk.
Professor James Mclaurin gave us an excellent overview of how Artificial Intelligence (AI) works and what it means for humans (and the planet) now and in the future. AI has been in existence for decades, however the emergence of Chat GPT earlier this year has showed how this technological advancement has and will disrupt many day-day activities and industries.
In the past AI was more focused on what is known as Narrow AI – where the application of AI is very limited in scope and task oriented, but the knowledge gained by the AI in this application is not automatically applied to other tasks. Siri, Alexa and Google Assistant are examples of Narrow AI.
General AI uses more complex datasets to solve a number of tasks using general human cognitive abilities, which all aim to do most things better than humans. such as chatbots, social media/music/video streaming recommendations (based on prior preferences), colour recognition, playing chess, self-driving vehicles.
Generative AI, which is a more complex type of predictive AI (what comes next) using large language models (LLM) and patterns, to create outputs in a wide range of applications.
James showed a video using Generative AI where he was speaking in Hindi on a topic, that he originally recorded speaking English, but the AI had used the sound of his voice (including inflections) and facial movements, and made the video look like he was fluently speaking in Hindi.
The use of AI has a very wide range of applications in many industries, and we are still at the early stages of use and regulation.
At universities students are learning how to better use AI given it will become an applied usage in their working careers beyond studying. Essays are becoming less common, and more physical presentations are being preferred, due to how good the AIs are at writing essays for students. Experiments are being developed much faster and more accurately with better results.
James recommended some great reading on the topic such as “The AI Revolution in Medicine” by Peter Lee et al where AI has delivered better patient outcomes when compared to a group of medical experts, and papers on “The Turing Trap” based on the how WWII codebreaker, Nile Turing’s views on machine intelligence which can either be used to make people more powerful but can also try to replicate (replace) people.
As generative AI is still relatively new, it is difficult to make bold predictions on the future, but James believed that it won’t necessarily replace all jobs, but will change the way roles are allocated and employment roles will change in response – things may get cheaper as a result. The new world of AI will provide new companions to how we do things a bit like how the Internet evolved from its early adoption to today, and that many general purpose robots using AI will become more commonplace.
Rotary International Convention, Singapore, 25-29 May 2024.
