How much control are we willing to hand over to artificial intelligence?

In Episode 55 of These Two Guys, George Henry and Larry Richardson dig into some unsettling recent developments involving AI — including a report that AI misidentified cargo aboard a Chinese freighter as being connected to a nuclear weapons program, raising questions about what happens when governments act on AI-generated intelligence.

Then there’s cybersecurity. The guys discuss researchers using Anthropic’s AI capabilities in security testing involving OpenAI, AI models operating outside expected cybersecurity boundaries, and what happens when powerful AI tools are placed in the hands of skilled hackers. The episode notes the OpenAI incident was ultimately an ethical security test, with the vulnerability reported and fixed — an important distinction.

George and Larry also get into the bigger question: What happens when AI becomes capable of finding solutions its creators never specifically instructed it to use? And where should the guardrails be when AI is involved in military, cybersecurity, surveillance and other high-stakes decisions?

Plus: Alexa listening, smart TVs, WarGames, Colossus: The Forbin Project, Eagle Eye, AI “cheating” to accomplish objectives — and the possibility of actually bringing Alexa on These Two Guys for an interview.

What do you think? Are the fears surrounding AI overblown, or are we giving these systems too much authority too quickly?

👍 Like the episode
💬 Tell us what you think in the comments
🔔 Subscribe for more These Two Guys
🌐 these2guys.com