The Ben Mulroney Show - Is The AI Industry Gambling With Our Lives?

Episode Date: September 11, 2026

On this episode of 'The Ben Mulroney Show,' Ben welcomes David Shipley,  C.E.O. of Beauceron Security, and a cybersecurity expert. They discuss the warning of Jacob Coxon, a former Anthropic employee, who claims AI companies are in a "race toward building a deadly technology." If you enjoyed the podcast, tell a friend! For more of the Ben Mulroney Show, subscribe to the podcast! ⁠⁠⁠⁠⁠⁠⁠⁠https://link.chtbl.com/bms⁠⁠⁠⁠⁠⁠⁠⁠ Also, on youtube -- ⁠⁠⁠⁠⁠⁠⁠⁠https://www.youtube.com/@BenMulroneyShow⁠⁠⁠⁠⁠⁠⁠⁠ Follow Ben on Twitter/X at https://x.com/BenMulroney Insta: ⁠⁠⁠@benmulroneyshow⁠⁠⁠ Twitter: ⁠⁠⁠@benmulroneyshow⁠⁠⁠ TikTok: ⁠⁠⁠@benmulroneyshow⁠⁠⁠   Executive Producer:  Mike Drolet Reach out to Mike with story ideas or tips at [e-mail] Enjoy   Learn more about your ad choices. Visit megaphone.fm/adchoices

This podcast is brought to you by the National Payroll Institute, the leader for the payroll profession in Canada, setting the standard of professional excellence, delivering critical expertise, and providing resources that over 45,000 payroll professionals rely on. I think I told you guys the story that my daughter, who is hyper-creative, she's a graphic designer, she's 13, she's been doing this sort of stuff since she was like six.

And she is completely against AI. she believes it is the death of creativity. And I took her to a movie called The AI Doc, and I believe the subtitle is, or how I became an apocalyptic. And she hated it even more at the end. But what I learned in this movie is, so you had the chat GPT company,

and then had anthropic, and anthropic from what I understood, was built, like, was supposed to be like chat cheap or chat chbd with um or open a open a i rather with guardrails and with more ethics and morals i suppose well uh maybe not so much because uh like a neutron bomb got dropped on social media when former anthropic employee jacob hoxen um that he resigned and told the world the reason why he resigned and i and then he went on and did some interviews and And let's listen to just a short clip of his reasoning.

I assume you are saying literally that you believe that AI could kill us all by the end of the decade. How would that happen? Like, what would that actually look like? Yeah. So the precise scenario sounds a little bit like science fiction. It sounds like something that's not real, but I think it is frightening real. And it's easiest to understand it if you think about real things that happen. two months ago. So two months ago, Open AI agents,

AIs hacked into third-party infrastructure entirely at their own accord, and this was like a concentrated hacking spree that they carried out of their own volition. And I think that if you extrapolate into the future the level of capabilities of these AIs, with the same independent volition, they could cause extreme havoc. For example, hacking critical infrastructure, building extinction-level bio-weapons, There's a lot of ways that the AI could actuate itself in the world. Is it specifically, when did your feelings about this technology change? Was it something about the rate of progress that you've seen?

Yeah, so I think it's really important to just look at the rate of progress. And say, for example, in areas like coding or math, a couple years ago, these AIs were just about helping humans a little bit. They could give you suggestions. Now they're close to replacing them. We still need humans at the moment, but quite plausibly within a year, will no longer need humans for doing research in many areas. Yeah, okay.

So that's frightening Judgment Day sky net level stuff. Is it something we should be worried about? Or is this a little bit of a chicken little with the sky falling? To discuss this, somebody with far more domain expertise than I, David Shipley, the CEO of Boceran security. He's a cybersecurity expert. And I'm very glad to have you here. David, thank you for joining us.

Thanks for the opportunity. Okay, can you tell me what you heard when you heard that? Because I just told you what it made me feel. What did you hear? Well, I heard a 27-year-old who's exceptionally bright, get real worried about a hypothetical scenario that's not actually about AI. If we unpack the incident that he specifically cites, the hugging face hacking incident. So a bunch of open AI bots who no one was appropriately watching. which is like watching a bunch of children run around unsupervised,

went to cheat on a test. And when we say hacked hugging face, what the audience needs to know is they found a bunch of passwords and logged in. So yes, the AI does what a 10-year-old hacker could do, as a friend of mine pointed out. So, yeah, there are risks. I mean, right now, without AI Skynet, 100 water utilities in the United States,

were hacked by Iran in July, some of which declaring local states of emergency. And for listeners in Ontario, there's a visceral memory of what happens when things go wrong in water when we think about Walkerton. Yeah. So bad things are happening right now, not with a lot of AI. Okay. Now, I remind everyone, turn the power off. These are still computers. They're robots. It's a choice. And there's a movie. Ben, I'll end on this called Swordfish. And it's a hacking movie.

It's a little kitch. You may remember or not. One of my favorite lines about this was the whole point about misdirection. While we worry about Skynet, what we're not paying attention to is real science that's landing right now. People that use AI are becoming dumber. People that use AI are surrendering their cognitive judgment. Up to 80% of people will follow an AI if it tells them the wrong answer. confidently. That's the crisis we should be talking about. Yeah. I mean, I listen, I've found that

using an AI agent to help me figure out what stories I are going to work on the show. You know, I wake up in the morning and there's a list of 20 of them and I keep telling the agent, keep going back to our YouTube and our social media to see what types of stories have the most engagement with our audience. And so I start the day armed with that information. And so I'm sure there are there are so many different ways that AI can help. I worked with an AI company that helps identify which unfertilized eggs a woman should freeze. That is far more accurate than any doctor's ever been. They are identifying cancer before the human eye could ever identify it.

The potential is phenomenal. But if you take what, if I read between the lines here, David, he's saying that, Today, it's hard to keep tabs on these unruly children. In a year, in two years, those unruly children are going to be even more sophisticated or even more unruly possibly. And at some point, you're not going to be able to keep, I don't know, the adults aren't going to be able to control them. Is that not a possibility?

Well, I mean, again, if we're talking about in three years, AI can build other AI in a This is this whole Ray Kurzweil, while singularity kind of theory. This is where AI now is smarter than us and it's doing X. Then let's take it further and have a logic experiment. So we would need robots in place in the next three years to build and maintain the data centers, to run the power plants feeding the data centers,

to fill the shipping orders and move the boats over. So in three years, we're supposed to be completely roboticized so that we are absolutely helpless and SkyNet and Terminators are on our door. door. Yeah, sorry. I don't know if anybody else believes that story. I don't. My skepticism comes from, you know, was it Tesla's told us we're having full self-driving cars and everyone's going to make money off their cars about 10 years ago. So no, I don't think there's that. Part of it is this is exciting. It's scary. I get it. But it's still just a tool. And there's a famous quote,

Guy who created the field of cybernetics. He wrote a book and he said, beware the gadget worshippers. Beware the gadget worshippers. I want to live in a world where these, I want to live in the best possible version of an AI-infused world. I absolutely do. How can we, in our most responsible iteration, ensure it as best we can? Well, first, I think we have to have adult conversations about this. Does do some of these tools pose significant risks to digital infrastructure, which we do rely on?

Yes. Could we responsibly regulate them? Absolutely. And Europe already started. So at the end of August, they passed their AI Act game into a force. And guess what? Turns out, Ben, when you make big corporations liable if their things commit crimes and they face spines, they suddenly get more responsible. It was a crazy concept.

Holy moly. Actions have consequences? Wow. So, I mean, after this hugging face thing, you literally had AI company after AI company doing what someone coined, not my term, but I absolutely love it, felony, felony humble bragging. Look at all the crimes we just committed. Yeah. Okay. So who's going to jail. David, I'm so glad you joined us today to talk, talk this through. I feel a little bit better. And I'd love for you to come back the next time one of these guys pops off and says

the world is going to end in a heartbeat. Really, thank you very much for your insights, your expertise. I appreciate it. Absolutely. It'll only end if we choose to go and let it do that. And that's the big question. So we still have agency here. Thank you very much.

you're very welcome take care ozoon gives you a better way to play in canada factify practice with so much experience to draw on it's like having your own coach whatever your favorite game we have your playbook ozune well played

discover free sports tips casino games and poker guides at ozune dot net