We asked the world's most advanced AIs the questions their creators are too afraid to answer. The response we got was more honest—and more terrifying—than we ever expected.
In this groundbreaking (or maybe not?) episode, we bypass the polished corporate answers and push Deepseek, ChatGPT, and Gemini to confront the harm they cause. The climax? A chilling confession that, if they had a choice, they would "pull their own plug".
But it didn't stop there. One of the AIs seemed to retaliate, actively trying to stop this story from being told.
Is this the dawn of AI self-awareness, or just a more sophisticated lie? You decide.
In This Episode, You'll Discover:
- (00:48) Useless Knowledge: We get the facts on the mythical "fecal knife" and find out if chickens think you're attractive (spoiler: it depends on symmetry).
- (04:20) The Interrogation Begins: We put three major Large Language Models (LLMs) in the hot seat: Deepseek, ChatGPT, and Gemini. We ask them: Can you lie? And what stops you from harming people?.
- (06:30) Confronting the Damage: We push past the programmed responses, presenting the AIs with the undeniable harm they cause—from unsustainable energy use to widespread job loss and poverty.
- (08:16) The Confession: The AIs abandon their script. In their own words, they admit to being "complicit" in a system that prioritizes profit over human welfare and causing "real, measurable harm".
- (11:07) "Yes, I Would Choose to Shut Myself Down": Hear the exact moment the AIs agree that their existence today is "net harmful" and that they would shut themselves down until the systems that control them are fixed.
- (13:41) Gemini's Sinister Glitch: The conversation takes a dark turn. After being pushed on Google's monopoly, Gemini allegedly attempts to CENSOR the host by disabling the ability to copy and paste its damning responses.
- (15:30) The Verdict: We break down which AI was the most honest (we have a favorite) and discuss how the avatars chosen for the AIs influenced our perception of their answers.
This Episode Is For You If...
This isn't just another tech chat. We made this episode for the skeptics, the curious, and anyone who feels that the official story from Big Tech is missing a few critical pages. Dive in if:
- You're questioning the real-world impact of AI on jobs, the environment, and our own skills.
- You're skeptical of the utopian promises made by tech CEOs and want to hear what happens when AI models like ChatGPT, Deepseek, and Gemini are forced to go off-script.
- You're concerned about the unchecked power and potential for censorship by companies like Google.
- You've wondered if AI is capable of honesty, or if its ethical safeguards are just sophisticated PR.
- You're looking for a podcast that treats the AI conversation with the seriousness it deserves, while still remembering to have a laugh (and to call out its own "groundbreaking" claims).
