What do tech pioneers think about the AI revolution? - The Engineers, BBC World Service

Share

Summary

Caroline Steel hosts 'The Engineers' at Imperial College London, focusing on the AI revolution. She introduces three world leaders in AI: Paolo Pirjanian, CEO of Embodied; David Silver, Principal Research Scientist at Google DeepMind; and Regina Barzilay, distinguished professor for AI and Health at MIT, who discuss their work, the impact of AI on humanity, and address audience questions on AI regulation, job displacement, and its role in human development.

Highlights

The AI Revolution and Its Pioneers
00:00:00

Caroline Steel welcomes listeners to 'The Engineers' at Imperial College London, focusing on the technical revolution of artificial intelligence. She introduces a panel of three world leaders in the field: Paolo Pirjanian, founder and CEO of Embodied, working on emotionally intelligent robots for child development; David Silver, principal research scientist at Google DeepMind, known for AlphaGo and working on artificial general intelligence; and Regina Barzilay, distinguished professor at MIT, renowned for her breakthroughs in early breast cancer detection and antibiotic discovery.

Regina Barzilay's Journey into AI in Oncology
00:02:10

Regina Barzilay shares her personal journey into applying AI to oncology after her breast cancer diagnosis in 2014. She realised the lack of AI in medical diagnostics and treatment, despite being treated at a top hospital and working at MIT. This experience motivated her to bridge the gap between advanced technology and patient care, leading to a collaboration with Dr. Connie Lehman to use AI for early cancer detection.

David Silver and Reinforcement Learning
00:03:48

David Silver discusses his path to AI through the games industry and his PhD in reinforcement learning. He explains reinforcement learning as a method where a system learns from experience and trial and error, similar to how humans and animals learn. He clarifies that for a machine, 'reward' is a numerical signal—a positive number for a desirable outcome and a negative one for an undesirable outcome—driving its learning process.

Paolo Pirjanian on Emotional Robots for Child Development
00:05:41

Paolo Pirjanian explains his motivation for creating emotionally intelligent robots, stemming from his own experiences of alienation in foreign countries. He highlights the growing prevalence of autism in children and how robots can serve as 'training wheels' to help them practice social skills like eye contact and turn-taking, fostering their integration into society. He clarifies that these robots are not meant to replace human contact but to facilitate it.

AI's Superiority in Cancer Detection and Drug Discovery
00:07:09

Regina Barzilay elaborates on AI's capability to understand cancer better than humans, particularly in dealing with uncertainty and predicting outcomes. She recounts how her own mammograms, re-evaluated by AI, revealed a tiny cancer two years earlier than human diagnosis, demonstrating AI's ability to process complex data and reduce diagnostic guesswork. She also discusses her team's success in using AI to discover a new antibiotic, effective against drug-resistant bacteria like E. coli and MRSA, highlighting AI's potential in drug discovery where economic factors deter pharmaceutical companies.

The Complexity of Go and Artificial General Intelligence
00:08:14

David Silver explains why Go was a greater challenge for AI than chess. Despite seemingly simple rules, Go possesses immense complexity, requiring intuition and creativity—traits traditionally considered uniquely human. He describes how AlphaGo had to imagine game outcomes hundreds of moves ahead, relying on an intuitive sense rather than just tactical lookahead. He then introduces Artificial General Intelligence (AGI), aiming to create systems that can approach any number of problems with human-like intelligence, not just single tasks.

Challenges and Future of AI in Medicine and Assistive Care
00:13:43

Regina Barzilay addresses why AI isn't more widely used in medicine, citing issues with regulation, billing models that disincentivize AI use (as faster treatment can mean less revenue for doctors), and the slow translation of technology into patient care. Paolo Pirjanian discusses the future of social emotional AI systems, particularly in assistive care for the elderly. He envisions robots as companions that can help with independent living and dignity, going beyond social interaction to include tasks like cooking and walking, predicting this could be a reality within the next decade.

AI and Human Culture: Creativity and Collaboration
00:16:22

David Silver talks about Google's Gemini, aiming to perform diverse tasks from tax returns to novel writing. He believes AI won't 'take over' human culture but will serve as a powerful tool for human creativity, citing examples like music authoring systems that accelerate songwriting. He envisions a future where AI and humans collaborate to produce enhanced creative works, like more amazing novels.

AI Regulation and Its Dilemmas
00:17:43

The panel addresses questions about AI regulation. David Silver supports regulation but stresses its need to be tailored to specific areas, like medicine versus chatbots, due to AI's diverse applications. Regina Barzilay expresses concern that excessive caution in regulation could hinder AI's potential to alleviate suffering from incurable diseases. Paolo Pirjanian highlights the conundrum of regulation, balancing risks with the strategic importance of AI; he believes effective regulation would require global international agreement, which history shows is difficult to achieve, and worries about an 'arms race' if some nations slow down development while others do not.

AI in Sports and Human Learning
00:21:32

David Silver discusses AI's potential to improve sports performance, citing Google DeepMind's collaboration with Liverpool Football Club to enhance tactics. A young audience member questions if humans will stop learning as AI develops. David envisions AI as a personal friend and teacher, fostering more learning. Regina believes AI will facilitate human focus on core ideas by handling mundane tasks, like language correction in writing. Paolo suggests AI will make humans more prolific, similar to how calculators aided past geniuses, enabling complex tasks in mere hours that traditionally took years.

Recently Summarized Articles

Loading...