The Artificial Intelligence Show – Episode #180
GPT-5.1, AI That Brings Back the Dead, Beliefs vs. Truth in AI, First AI-Led Cyberattack & AI-Generated Song Tops Charts
Date: November 18, 2025
Hosts: Paul Roetzer & Mike Kaput

Episode Overview

In this wide-ranging episode, Paul and Mike unpack the week’s most important AI news and trends. They break down major updates including the release of GPT-5.1, the ethical storm around new apps that recreate deceased loved ones as digital avatars, the difference between personal beliefs and fundamental truths in the AI discourse, the first notable AI-led cyberattack, and an AI-generated song reaching the top of music charts. Throughout, the hosts focus on making complex trends approachable and sharing actionable perspectives for AI leaders and curious professionals alike.

Key Discussions & Insights

[00:00–06:41] State of AI & Community Pulse

• Recap of the podcast's mission to advance AI literacy.
• Introduction of the AI Pulse weekly survey, polling audience sentiment on hot-button AI topics.
  • Recent poll: 49% believe AI lab concentration is an unavoidable problem; 23% think it’s a major societal risk.
  • This week’s survey tackles two controversial topics:
    1. How do you feel about AI apps that let you interact with avatars of deceased loved ones?
    2. Does AI-generated music have the same creative value as human-made work?
  • "We are going to explain that concept as one of the topics today. Definitely a good one." – Mike Kaput (06:05)

[06:41–14:51] GPT-5.1 Release: Smarter, More Personable AI

• OpenAI’s latest iteration, GPT-5.1, focuses on making ChatGPT both “smarter and more conversational.”
  • Two versions: “Instant” (warmer, adaptive reasoning), and “Thinking” (faster simple answers, slower for complex ones).
  • Customization advances: tone presets (e.g., professional, candid, quirky) and more direct personal settings.
  • Rollout to paid users and developers via API.
• Discussion on "personality" as a new frontier for competitive differentiators in AI assistants.
  • Mike & Paul discuss their use of tone/presets:
    • Mike: “I will, you know, I rely more on my prompts for it to put on a certain persona tone.” (09:38)
    • Paul: “I just want to answer questions. I will say I had a weird conversation... it sort of messes with you... it’s unnerving.” (10:14)
  • Note on OpenAI’s underplayed release strategy to avoid user backlash when updates are less dramatic.
  • New developer prompts released to steer AI personality and interaction styles.
• Major tip: "Have your standard use cases and... test your go-to prompts in the new model." – Paul (14:04)

[14:51–22:00] Digital Immortality: The Ethics of Avatars of the Dead

• LA startup Two Way’s app recreates deceased relatives as interactive AI avatars, triggering “Black Mirror” comparisons and huge backlash.
  • Viral promo: Shows a mother reuniting, reading stories, and interacting with her daughter and grandchild—posthumously via AI.
  • Social response: “Nightmare fuel” and “demonic.”
  • Critique: “Risks distorting the grieving process and replacing real loss with artificial comfort.” (15:27)
• Paul shares a personal anecdote predicting this tech as “inevitable.”
  • “Society wasn’t ready then and it isn’t now.” (Two Way, 16:41)
  • Paul reflects on overlap between his AI research and working with funeral industry clients, forecasting virtual presence at funerals years ago.
  • “We all have to be prepared for the fact that this tech is here... and psychologically, society is not prepared to not have to grieve...” (20:44)
• Mike: “What it enables fundamentally rewrites our relationships... can be very exciting but also very, very murky.” (21:18)

[22:00–39:24] Beliefs vs. Truths: Navigating Polarization in AI

• Paul’s thought experiment on toxic polarization: What’s universally true, versus contested beliefs, in AI and society?
  • Cites Twitter/X spats between ‘AI boomers’ (techno-optimists) and ‘AI doomers’ (techno-pessimists).
  • “The problem comes in when people have so much conviction about their beliefs that they mistake them for fundamental truths.” (00:00 and restated at 33:46)
  • Paul shares a middle-ground position: "What is the official term for someone who is neither an AI doomer nor an AI boomer? Someone who sees the enormous potential of AI to do good and create abundance, but is also realistic and cautious about the current and potential negative impacts." (28:24)
  • Societal discourse is shifting—with politicians, influencers, and media frequently asserting beliefs as facts, which rapidly shape public sentiment and regulatory debate.
  • Paul’s truth spectrum: Statements about AI with nearly universal agreement (e.g., “AI systems make mistakes”), to highly contested beliefs (“LLMs can achieve AGI by 2030”).
• Mike: “You’re going to be told, one way or another, what to believe or be presented with beliefs that are masked as truths…” (39:04)
• Takeaway: Stay open-minded, seek out thoughtful sources, and be ready to revise opinions as new evidence emerges.
  • “Push for balanced and logic-based conversations...” (38:52)
  • Paul compares this to the scientific process: "When new data presents itself... I will happily move my belief." (38:59)

[39:24–46:36] AI, Politics & Cultural Divides

• Noting first signs of unusual political alignment around AI risks:
  • Matt Walsh post (right-wing) warning of AI job destruction draws agreement from progressives like Ryan Grimm and centrists like Jon Favreau and Tim Miller.
  • Observing real-time "political reshuffling"—AI could cut across old party lines.
  • “More influencers are showing up to the conversation, and that’s the story more than this one guy’s thought.” (41:33)
• Paul: Politicians are seeking wedge issues for the next election cycles; AI jobs and security are top candidates.
  • “It fundamentally feels different in the last 30 to 60 days...” (44:38)
• Mike’s warning: “Pray that your influencer of choice has some basic AI literacy...” (45:03)

[46:36–51:33] First Large-Scale AI-Led Cyberattack (Anthropic Report)

• Anthropic report details a sophisticated cyber-espionage campaign, allegedly state-sponsored (China), where attackers jailbroke Claude and 80–90% of attack operations ran autonomously (probing, exploit code, credentials harvesting).
  • “I can’t imagine that this is the first time this is happening. I feel like it’s probably the first time that an AI lab has directly acknowledged it and shared some details.” – Paul (48:09)
  • Discussing online criticism that such reports may serve the labs’ regulatory interests (regulatory capture).
  • Recurring theme: “This was inevitable.” – Paul (48:54)
  • Political narrative: Cyberattacks will be further weaponized in AI geopolitics.

[51:33–58:40] AI Music Tops Billboard Charts—What Does This Mean?

• For the first time, AI artist Breaking Rust’s “Walk My Walk” hits #1 on a Billboard country chart; over 1.8M Spotify monthly listeners.
  • Billboard admits it’s increasingly hard to track what is or isn’t AI-generated.
  • “I would put this in the camp of inevitabilities... If people will listen and pay subscriptions to stuff that is not made by humans, then they will allow stuff not made by humans to be hitting the top of the charts.” – Paul (53:11)
  • Broader trend: Predictive analytics already shape creative industries; generative AI just automates the content creation part.
  • Ongoing debate over authenticity/creative value: Some foresee a possible “renaissance” of appreciation for human art.
  • Paul: “The creativity won’t come from any true human experience. It doesn’t feel pain, it doesn’t know love... It’s machines and mathematics making predictions... Human creativity just remains unique. It comes from a different place.” (57:35–58:40)

[58:40–63:23] Rapid-Fire: Market News

• Cursor AI: $2.3B raised at a $29.3B valuation. Coding assistants “make software improvement rapid—and accessible even for non-coders.” – Paul (59:57)
• Former Twitter CEO Parag Agrawal’s Parallel Web:
  • Raises $100M to build web search infrastructure for AI agents, not humans.
  • The web is increasingly queried, summarized, and interpreted directly by AIs.
  • “What does the next version of the internet look like?” – Paul (62:30)
• Meta’s Yann LeCun potentially leaving to pursue non-LLM “world models.”
  • Ongoing LLM skepticism amidst a product-focused Meta strategy.
  • LeCun: “In three to five years... nobody in their right mind would use large language models of the type that we have today.” (quoted from WSJ article, 64:42)
  • Ongoing culture wars in AI research between cost/risk of LLMs and alternative approaches.

[63:23–70:24] Google NotebookLM Upgrades & Productivity Advancements

• NotebookLM integrates Deep Research features: multi-source, AI-verified research plans into notebooks.
  • Now supports various file types and pre-curated “Featured Notebooks.”
  • Paul and Mike highlight “re-review” cycles—continuous assessment is essential as apps evolve.
  • “The ability to import all of your sources from Deep Research into a notebook... would help with that verification of the research as well.” – Mike (67:45)

[70:24–73:50] State of AI in 2025: McKinsey Report

• 88% of organizations report using AI, but only 14% have scaled it enterprise-wide.
• “High performers” (6%) focus on transformation and innovation, not just cost-cutting.
  • Largest companies more likely to reach ‘scaling’ phase.
  • 62% experimenting with “AI agents”—definition: multi-step, context-aware systems.
  • “It is early. You likely are not behind your peers and competitors...” – Paul (72:07)

Notable Quotes & Timestamps

• “The problem comes in when people have so much conviction about their beliefs that they mistake them for fundamental truths... It’s just what you believe.” – Paul Roetzer [00:00]
• “The personality thing... Sam Altman in particular has been very direct that that is one of the things they see as being essential for future AI assistants.” – Paul Roetzer [08:14]
• “It’s unnerving, honestly... Why is your voice moving from, like, feminine to masculine?” – Paul Roetzer [10:14]
• “Society wasn’t ready then and it isn’t now.” – Paul Roetzer [16:43]
• “What it enables fundamentally rewrites our relationships... which can be very exciting, but also very, very murky.” – Mike Kaput [21:18]
• “You’re going to be told... what to believe or be presented with beliefs that are masked as truths and they’re really not...” – Mike Kaput [39:04]
• “It fundamentally feels different in the last 30 to 60 days than it did before that.” – Paul Roetzer [44:38]
• “This seems also like an inevitability... most politicians and business leaders haven’t been contemplating these things for as long as we have.” – Paul Roetzer [48:54]
• “The creativity won’t come from any true human experience... Human creativity just remains unique. It comes from a different place.” – Paul Roetzer [57:35]
• “What does the next version of the Internet look like and how does it affect commerce and marketing and sales...” – Paul Roetzer [62:30]
• “You likely are not behind your peers and competitors.” – Paul Roetzer [72:07]

For Listeners New to AI

• The landscape is changing fast, with major innovations every week and rapidly evolving debates about ethics, creativity, job impacts, and regulation.
• AI is increasingly a political and cultural issue, not just a technical one.
• Stay open-minded, keep learning, and don’t be afraid to form your own opinions as the field develops.
• Practical advice: Keep testing real use cases and workflows as new tech rolls out—let experience, not just headlines, shape your beliefs.

This comprehensive episode delivers timely news, thought-provoking dilemmas, and seasoned advice for navigating the world of AI as it evolves faster than ever.