Jamie Sherrah
Dr. Jamie Sherrah has over 25 years experience in machine learning, computer vision and commercial software development. He is Director for Defence at the Australian Institute for Machine Learning (AIML), and managing director of the machine learning software company Inject AI. At AIML Jamie has been involved with the Art Intelligence programme, including such projects as a singing synthesiser.
Website LinkedIn Twitter
My favourite AI in culture project at the moment is… natural language generation models such as GPT-2. These gigantic models have condensed millions of web pages into a single generative model that can draw on the breadth of human culture to automatically write convincing text.
By 2030 I would like to have an AI app which… is my broker to all of the internet, filters information and proactively finds me interesting stuff and connects me to interesting people.
My worst AI nightmare is… one of the visions of the future in Black Mirror, like the episode where an AI-driven brain implant drives its wearer crazy or a robot hunting a human.
The future of AI needs… theory-of-mind built into machine learning so that computers and humans can communicate more effectively.
Jamie Sherrah's contribution: "What are AI-driven music recommendation tools doing to us?"
Website LinkedIn Twitter
My favourite AI in culture project at the moment is… natural language generation models such as GPT-2. These gigantic models have condensed millions of web pages into a single generative model that can draw on the breadth of human culture to automatically write convincing text.
By 2030 I would like to have an AI app which… is my broker to all of the internet, filters information and proactively finds me interesting stuff and connects me to interesting people.
My worst AI nightmare is… one of the visions of the future in Black Mirror, like the episode where an AI-driven brain implant drives its wearer crazy or a robot hunting a human.
The future of AI needs… theory-of-mind built into machine learning so that computers and humans can communicate more effectively.
Jamie Sherrah's contribution: "What are AI-driven music recommendation tools doing to us?"