AI Avengers

Self-Aware or Just Seemingly Conscious AI?

58 min · 29. aug. 2025
episode Self-Aware or Just Seemingly Conscious AI? cover

Description

Is AI sentient, or just staggeringly convincing? In this episode we unpack why so many people feel like their chatbot is alive, riff on Mustafa Suleyman’s “seemingly conscious AI,” and test the line between personality, persuasion, and genuine awareness. From Vision’s Mind Stone to The Matrix and TRON, we use pop culture to ground a serious debate: belief vs. reality, and how easily both can be manipulated at scale. Then we zoom out to the real-world stakes, commercial arms races, safety guardrails that don’t kill speed, and why defenders may need un-guardrailed “agents of doom” for honest red teaming even as we keep values front and center. We close with the hard stuff: privacy (shared chats, searchable links, wearables that capture everything), OSINT made trivial by multimodal models, job displacement vs. new work, and candid bets on superintelligence timelines. It’s a reality check, a caution flag, and a practical playbook for leaders who have to steer now.

Comments

0

Be the first to comment

Sign up now and become a member of the AI Avengers community!

Get Started

1 month for 9 kr.

Then 99 kr. / month · Cancel anytime.

  • Podcasts kun på Podimo
  • 20 lydbogstimer pr. måned
  • Gratis podcasts

All episodes

2 episodes

episode Self-Aware or Just Seemingly Conscious AI? artwork

Self-Aware or Just Seemingly Conscious AI?

Is AI sentient, or just staggeringly convincing? In this episode we unpack why so many people feel like their chatbot is alive, riff on Mustafa Suleyman’s “seemingly conscious AI,” and test the line between personality, persuasion, and genuine awareness. From Vision’s Mind Stone to The Matrix and TRON, we use pop culture to ground a serious debate: belief vs. reality, and how easily both can be manipulated at scale. Then we zoom out to the real-world stakes, commercial arms races, safety guardrails that don’t kill speed, and why defenders may need un-guardrailed “agents of doom” for honest red teaming even as we keep values front and center. We close with the hard stuff: privacy (shared chats, searchable links, wearables that capture everything), OSINT made trivial by multimodal models, job displacement vs. new work, and candid bets on superintelligence timelines. It’s a reality check, a caution flag, and a practical playbook for leaders who have to steer now.

29. aug. 202558 min
episode Me, Myself, and AI: Cheat-GPT & the Tinkerer’s Code artwork

Me, Myself, and AI: Cheat-GPT & the Tinkerer’s Code

Two builders pull back the curtain on how AI actually becomes useful, not in slides, but in real life. We swap origin stories (from Terminator-era dreams to Commandant-class copilots), then share the “cheat codes” that turn curiosity into capability: tinker first, screenshot errors to your model, iterate fast, and treat frustration as proof you’re learning. We wrestle with school bans and enterprise “do not pass go” edicts, outline sane guardrails that won’t strangle velocity, and tackle the fear cycle, deepfakes, job loss, and the flood of AI-generated content, by arguing for long-form, human storytelling and AI as accelerant, not replacement. Along the way: MCP/A2A agents, why passion beats pedigree in hiring, Lucasfilm’s playbook for adopting new tech, and how to become the AI-powered human your team can’t live without. This is an on-ramp, a reality check, and a call to learn in public—because the future favors people who ship

18. aug. 20251 h 0 min