The Problem With The AI Apocalypse: Connor Leahy

The Problem With...

James Smith

Society & Culture

4.9 • 9.5K Ratings

🗓️ 9 December 2025

⏱️ 120 minutes

🔗️ Recording | iTunes | RSS

🧾️ Download transcript

Summary

Check your testosterone levels: https://www.manual.co/smith I help small businesses make money online: https://www.jamessmith.business Try Neutonic: https://www.neutonic.com/jamessmith Free trial: Online coaching app: https://www.affordableonlinecoaching.com AI researcher and ControlAI founder Connor Leahy joins James Smith to confront the problem with artificial intelligence - from misinformation and manipulation to the terrifying speed at which machines are learning to outthink their creators. Connor, one of the world’s leading voices in AI safety, explains how we’re building systems that are more intelligent than humans…and why we don’t really understand how they work. They discuss the future of AGI (Artificial General Intelligence) and ASI (Artificial Superintelligence), exploring what happens when machines evolve faster than we can control them. Connor breaks down the moral, political, and psychological implications of AI, from fake online identities and emotional manipulation to the emerging arms race between nations and corporations desperate to be first. He explains: ◼️ Why AI is already outpacing human understanding ◼️ How propaganda and algorithms shape reality ◼️ The dangers of AI “boyfriends” and emotional manipulation ◼️ Why superintelligence could end human control ◼️ What governments and people must do before it’s too late Chapters 0:00 – Why the Future of AI Looks Dangerous 0:31 – What ASI Really Means for Humanity 1:29 – The Core Problem With Modern AI Systems 3:02 – How AI Is Built and Why We Can’t Control It 6:04 – The Hidden Dangers of Social Media + AI Manipulation 10:09 – AI vs AGI vs ASI Explained Simply 17:01 – How Fast AGI Could Turn Into Superintelligence 22:22 – Utopian vs Dystopian Futures With AI 29:02 – AI Relationships, Loneliness & Digital Partners 47:18 – Why AI Could Replace Human Connection & Autonomy Welcome to "The Problem With" where each week we look into a problem to get a better understanding of it. This podcast has no sponsors, only my businesses and investments. Please check out the links below. I'm on a mission to help men check their levels of Testosterone, more info here: https://www.manual.co/smith I help small businesses make money online: https://www.jamessmith.business Please check out and try Neutonic here: https://www.neutonic.com/jamessmith For a free trial of my online personal training app go here: https://www.affordableonlinecoaching.com Learn more about your ad choices. Visit megaphone.fm/adchoices

Transcript

Click on a timestamp to play from that location

0:00.0	I don't think the future looks good.
0:01.3	I don't think humanity is going to survive this by default.
0:03.3	And people sometimes ask me, like, surely humans would stop, you know, AI. And I'm like, really? Really? Have you take a look around how people are reacting to AI and how AI is manipulating people? Just wait until people start advocating for AI rights. Like, it's already starting, right, where people are like, oh, my girlfriend should have human rights. Like, she's real. Honor Leahy is the CEO of Conjecture.
0:22.0	An AI startup working on aligning AI systems to human values.
0:25.6	James and Connor discuss the weird developments in AI relationships, the scary future of AI.
0:31.6	And...
0:32.6	Connor, what is the problem with AI?
0:33.6	ASI would be a system.
0:35.6	It doesn't have to be a single model, which could be many models working together, that is smarter than all of humanity put together. Such a system, if it existed, it would be game over for human and humanity. And ASI could keep humans around if it wanted to. It could exterminate everyone if it wanted. This only gets worse. What is the solution? What can we do? What are the actionable points? points. What's the light at the end of the tunnel? The good news is that we haven't get lost. The bad news is that we're on track. So the main thing we must do is... This podcast is brought to you by Newtonic. Look, we don't have any fancy sponsors to this podcast. so thank you to all of you that drink Newtonic.
1:12.3	It helps keep this podcast going.
1:14.3	If you haven't tried it, bloody suggests that you do.
1:14.9	It's really good.
1:24.5	It's like an energy drink, but much better, although the marketing department say I'm not really allowed to say that. And if you're not into caffeine, we have pills and powders that taste amazing and work really well. I could go on about all the benefits, but why don't me just go back to the podcast episode?
1:25.4	That'd be nice, wouldn't it?
1:26.4	Head to Newtonic.com.
1:26.8	Bye-bye.
1:28.9	Connor, what is the problem with AI?
1:28.8	Well, starting with easy nice, wouldn't it? Head to Newtonic.com. Bye-bye. Connor, what is the problem with AI?
1:31.1	Well, starting with the easy questions, aren't we? AI is a really big topic. I mean, it's the thing that all of us is talking about. It seems to be everywhere, right? Like, you know, you got your chat, GPTs, you got deep fakes, you got full-on musical artists that are just AI generated now.
1:44.6	It seems like everything's being AI generated.
1:46.6	And this can make it kind of... deepfakes, you got full-on musical artists that are just AI generated now.
1:44.7	It seems like everything's being AI generated.
	...

Please login to see the full transcript.

Previous episode | Next episode

Disclaimer: The podcast and artwork embedded on this page are from James Smith, and are the property of its owner and not affiliated with or endorsed by Tapesearch.

Generated transcripts are the property of James Smith and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.