This podcast predicts the future of artificial intelligence

(From left) Dr Geoffrey Hinton and Nayeema Raza in conversation. PIC COURTESY/YOUTUBE

We were all wrong," admits Dr Geoffrey Hinton in conversation with podcaster Nayeema Raza. It's not every day that you hear this admission from a Nobel Prize awardee. Long before he became the Nobel Prize Laureate in Physics for his work on machine learning in 2024, Hinton joined Google Brain in 2012 (the AI arm of Google). He would later resign in 2023 citing the growing risks of AI for humanity. Hence, his admission in the podcast: That he, along with many like him, were wrong in predicting how soon Artificial Intelligence (AI) would become a tangible threat.

Throughout the 30-minute podcast, Raza throws oversimplified questions at the Nobel Laureate. Understandably so - the podcast titled Smart Girl Dumb Questions pitches questions that many of us have pondered upon, but never asked out loud. "Is there anything humans can do better than AI?" she asks. Hinton admits, "Absolutely nothing." To its credit, the podcast steers clear of conspiracy theories (except a short detour to take shots at Tesla and X owner Elon Musk's vested interests).

That begs the bigger question, what can AI do to us in 2026? Hinton lays out two major threats. Warfare, he says, will be starkly different. AI models will dictate whom and when to attack, leaving little room for emotions in war. Orchestrating cyberattacks, on the other hand, will become as easy as clicking a button. The biggest concern, he believes, is AI growing apathetic. "AI wants two things: To know more, and to have more control," Hinton notes. "At one point when your child is struggling to tie his shoelaces, you'd lose patience and take over. That's where we're at with AI. Control might soon change hands," he reveals.

The expert's solution to the quandary is unique: Teach AI to develop maternal instincts. Injecting empathy into the technology, he believes, is the only way it will care for humans when it eventually takes the wheel. Hinton admits he does not know the ‘how' behind it yet. While we wish the podcaster and the guest took a breather to focus on the positive impact AI could have in areas like healthcare, education, and STEM research, we'll cut them some slack given Dr Hinton's long standing stance: "We must act urgently, and act loud in the face of AI."

There may be some good news though, the guest hints. "No superpower in world politics wants another to become an AI superpower. In that sense, whenever one of them figures out how to disarm AI, they'll surely announce it to the world," he laughs. Raza plays a good host, creating enough room for the academician to walk in and out of humorous detours with ease.

Raza ends the podcast with a question that has the internet divided: Should you say thanks and please to your chatbot, lest you end up in its bad books once it figures out world domination? "Even if you're rude to your personal AI assistant, it will be smart, and more importantly benevolent enough to know you didn't mean it that way," he signs off. Omnipresent, all-knowing, and now supposedly a benevolent figure that sees us as its children - that rings a bell.

LOG ON TO Smart Girl Dumb Questions on YouTube