IBM Australia has launched a new podcast series aimed at helping listeners understand generative AI.
The podcast series, titled "IBM Uncharted: Generative AI Explored," is designed for executives and those interested in artificial intelligence, providing insights into the opportunities and challenges associated with generative AI.
Hosted by Miki Luong, Chief Marketing and Communications Officer at IBM, the podcast's first episode highlights discussions with AI experts and thought leaders from IBM.
"Generative AI has taken off so suddenly that it's easy to feel left behind," said Ms Luong. "When we hosted our Uncharted event series across Australia, exploring generative AI for an executive audience, we found that even the most accomplished leaders felt unsure about the full implications of this space. And if experts in their field don't feel confident leading a technology as groundbreaking as generative AI, then where does that leave the rest of us?"
Ms Luong continued, stating, "IBM Uncharted: Generative AI Explored is our way of trying to answer and amplify the many questions and conversations happening right within organisations. We hope our series will help listeners discover how artificial intelligence will continue to impact the way we work and live."
The inaugural episode features a conversation between Ms Luong and Adam Makarucha, Client Engineering Leader at IBM, where they delve into the basics of generative AI, its applications, and current limitations. The episode offers practical insights and examples of how professionals might use generative AI in organisational contexts.
Dr Makarucha explained the nature of AI intelligence, saying, "AI is not intelligent in the same way we think of intelligence." He distinguished between "narrow AI," which is data-driven and task-focused, and "general AI," which utilizes multiple inputs to reason similarly to humans, pointing out that general AI is still far from reality.
"Traditional AI is very prescriptive," added Dr Makarucha. "You know exactly what data it's trained on. It could be your insurance case, an interaction record, how much your policy is worth, data about you as a person. With gen AI, the input is a text box so it's as expressive as we want it to be. We can really go ahead and use it in any way we want."
In the podcast, Dr Makarucha elaborated on generative AI being a subset of narrow AI, capable of generating outputs from text, speech, and images through "self-supervised learning," continually refining its reasoning until accurate results are achieved.
The subsequent episode features Luong, Kaaren Koomen, Director of Government and Regulatory Affairs at IBM Australia and New Zealand, and Richie Paul, Partner and Generative AI Leader at IBM Consulting. They discuss the importance of establishing robust governance and ethical frameworks to ensure generative AI's successful long-term application, and they examine potential conflicts between corporate incentives and societal impacts.
In the third episode, Makarucha interviews Ajay Singh, Managing Director of Stretch 365, focusing on AI governance's vital role, examining real-world oversight failures, and addressing emerging threats such as prompt hacking.
The podcast series is set to run on a monthly basis, providing ongoing content about generative AI and its broader implications.
