Podcast

The CERN Sparks! Podcast - Future Intelligence

Can artificial intelligence be curious? Can it be creative? Can we create it to be ethical and push past the limits that exist today to imagine a world made better with AI in it? At CERN, machine learning already plays a pivotal role in the quest for advancing human knowledge through particle physics – from detector design to data acquisition and analysis. But can AI one day take over the role of scientists? Join hosts Abha and Mark on the Sparks! podcast as they journey through a series of conversations with leading minds shaping the realm of Future Intelligence.

 

Episodes

# Trailer - The CERN Sparks! Podcast

Artificial intelligence is transforming our world. Hear the sparks fly as Mark Rayner and Abha Eli Phoboo collide pairs of the leading coders, neuroscientists, policymakers, philosophers, psychologists and physicists who are shaping the future.

Then join us for the first edition of the Sparks! Serendipity Forum at CERN in September.

#1 Brainy AI — with Stuart Russell and Tomaso Poggio

Do we need to understand the brain to make progress in artificial intelligence? In the first podcast in the series, Stuart Russell and Tomaso Poggio contrast “deep learning” with our own organic neural networks. In an age of great demonstrations by the likes of Deep Mind and OpenAI, our guests make the case for focusing on controlled experimentation, and question the wisdom of using AI in science before it is fully understood. The mystery of intelligence, says Poggio, is the greatest problem in science today — if we solve it, we solve all other problems too.

#2 Quantum AI — with Maria Spiropulu & Vivienne Ming

Episode 2 collides two rockstars of the world of artificial intelligence to reimagine the field for the next generation. Is consciousness quantum or just me talking to myself? Could quantum computing unlock a step change in artificial intelligence? Our guests also get down to earth on the need for AI to tackle real-world data-poor problems from hiring bias to diagnosing manic episodes in bipolar sufferers. There is a recurring flaw in applied artificial intelligence, argues Ming. Machine learning is not a Deus ex machina for your company’s problems: expertise is queen, and innovation by gender and ethnic minorities is problematically undervalued.

#3 Creative AI — with Anima Anandkumar and John Ellis

We can dream, we can hallucinate, we can create — so how do we build those capabilities into AI? Deep-learning expert Anima Anandkumar and distinguished theoretical physicist John Ellis discuss the potential for artificial intelligence to one day collaborate with us in attacking the biggest unanswered questions in physics — questions which have outwitted humans for years. In a conversation ranging from the quantum nature of subatomic reality to the distributed intelligence of the octopus, our guests explore how AI might one day tackle questions which are conceptually boundless and infinite. “This would be truly stealing the theoretical physicists’ lunch,” says Ellis.

#4 Experimental AI — with Maurizio Pierini and Michael Doser

Particle physics is at a moment of truth. The discovery in 2012 of the Higgs boson promises to reveal a rich new structure for the vacuum and rewrite the history of the early universe, but a long list of fundamental questions remains, and physicists are faced with an awesome data flow from the Large Hadron Collider. In Episode 4, CERN’s Maurizio Pierini and Michael Doser explore using “unsupervised learning” to reveal nature’s mysteries. I’m really super excited about the next LHC run, says Pierini, because this is when we’re going to try these things for real.

#5 Ethical AI — with Nyalleng Moorosi and S. Matthew Liao

“We always had privacy violation, we had people being blamed falsely for crimes they didn’t do, we had mis-diagnostics, we also had false news, but what AI has done is amplify all this, and make it bigger,” says Google’s Nyalleng Moorosi. In Episode 5, she and philosopher S. Matthew Liao debate the delicate balance between personal moral agency, human rights and corporate responsibility in the brave new world of artificial intelligence. We need to understand more about these principles, not just to list them, says Liao, because then there’s a worry that we’re just doing ethics washing — they sound good but they don’t have any bite.

#6 Fast and slow AI — with Francesca Rossi and Daniel Kahneman

Francesca Rossi is an influential global leader in AI research. Daniel Kahneman is one of the greatest living cognitive psychologists. In the final podcast in the series, our guests take Daniel’s revolutionary “fast and slow” systems of thought as inspiration for rewriting AI, and debate the nature of thought itself. “I really find it difficult to imagine why there should be anything at which humans are essential in the domain of intelligence,” says Kahneman. Is there anything that humans can do that AI cannot in principle do?

 

Hosts