The KMO Show (KMO)

KMO talks with Pi, the most advanced AI chatbot he has encountered to date.

Bryan Cheong, a machine learning engineer, joins KMO for a mostly non-technical discussion of large language models, reinforcement learning from human feedback, the democratizing potential of the current moment in artificial intelligence, and why your shouldn't get attached to a centrally-controlled AI companion when you can make and run your own decentralized version.

KMO speaks with data scientist Danielle Boccelli about the possibilities and limitations of learning strictly from language. The topic is large language models, but this conversation was recorded before the announced release of GPT-4. So much of what we know and what we are is not encoded in language, so there's only so far AI can progress without learning from something other than just text scraped from the internet.

Episode Title:

AI and Large Language Models: Bridging the Gap between Public Discourse and Research Progress

Episode Keywords:

Artificial Intelligence, Large Language Models, AI Research, Transformer Technology, Text Generation, AI in Education, AI Dilemma, GPT-4, AI Applications, AI Limitations, Public Discourse, Embodiment, Content Creation, Contract Writing, Textural Artifacts, Commercial Pressures, Chatbots, Fact-Checking, Polanyi's Paradox, Artificial General Intelligence

Episode Tweet:

In this episode of the KMO show, join us as we explore the fascinating world of Artificial Intelligence and Language Models with Danielle Boachelli. Get ready to gain valuable insights into AI and its implications! #AI #languageModels #KMOshow

Episode Summary:

In this episode of the KMO show, I had an insightful conversation with Danielle Boccelli, a data scientist and PhD student, about artificial intelligence and large language models. We discussed the gap between public discourse and actual research progress, the limits of learning from text, and the potential applications and limitations of large language models and transformer technology. We also explored the complexities of text generation, the impact of AI on education, and the commercial pressures driving the development of AI language models. Finally, we touched on the AI dilemma presented by Tristan Harris and Aza Raskin, and the evolving capabilities of large language models like GPT-4. Join us as we delve into these fascinating topics and gain valuable insights into the world of AI and language models.

Chapters:

(0:00:21) - The Reality of AI

(0:07:14) - Limits of Learning From Text Alone

(0:18:21) - Limitations of Transformer-Based AI Technology

(0:25:07) - Exploring the Complexities of Text Generation

(0:31:01) - The Impact of AI on Education

(0:42:52) - Language Models and Consequences

(0:57:03) - Exploring AI Dilemma

Chapter Summaries:

(0:00:21) - The Reality of AI (7 Minutes)

Full show notes at: https://www.patreon.com/posts/81435006

There has always been a psychological element to warfare but in the 21st century the tactics used by military geniuses to control the behavior of their adversaries have now been coopted by governments, corporations, ideologies, and the general public in order to wage psychological warfare against you. Kaleb Gorman, author of Psychwars: Self-Defence Against Psyops, Propaganda, and Mind Control, joins KMO to discuss why we are susceptible to such manipulation and how best to resist it.

This wide-ranging conversation between KMO (host) and David Roman (historian and writer) covers several major topics, starting with Roman's recent trip to China and observations about its development over his 27 years of visiting the country. The discussion then delves deeply into science fiction literature, publishing, and the current state of the genre. The conversation also explores generational ships in science fiction, AI, cultural representation in fiction, and ends with a discussion about philosophy, particularly continental philosophy and the work of Slavoj Žižek.

Table of Contents:

1. Introduction and China Discussion (00:00-05:00)
   • David Roman's background in China
   • Observations on China's development
   • Contrasts between modernization and traditional elements
2. Technology and Development in China (05:00-10:00)
   • Discussion of AI and robotics advancement
   • Reality vs. social media representation
   • Infrastructure and development contrasts
3. Media Platforms and Content Creation (10:00-25:00)
   • Experiences with YouTube, Facebook, and Substack
   • Discussion of platform censorship and monetization
   • Content creator challenges
4. Science Fiction Literature (25:00-45:00)
   • Discussion of various authors including Peter Watts
   • Analysis of current trends in science fiction
   • Generation ship novels and themes
5. Cultural Representation in Science Fiction (45:00-60:00)
   • Discussion of diversity in future scenarios
   • Critique of current publishing trends
   • Cultural authenticity in character writing
6. Publishing Industry Analysis (60:00-75:00)
   • Changes in editing and publishing
   • Self-publishing vs. traditional publishing
   • Impact on science fiction genre
7. Philosophy Discussion (75:00-end)
   • Continental vs. Analytic philosophy
   • Discussion of Slavoj Žižek's work
   • Roman's work on Chinese and Western philosophy

In this episode of The KMO Show, host KMO speaks with Kevin Lynn, Executive Director of the Institute for Sound Public Policy, about the 2024 election results and broader historical patterns. The conversation begins with an AI-generated introduction acknowledging their shared history in the Peak Oil community and mutual interest in Rudyard Lynch's analysis of historical crisis patterns.

Kevin brings unique perspective from his varied career - from Army officer to accounting executive to political organizer. The discussion explores several key themes:

• The parallels between current events and historical periods like the Thirty Years War
• The emergence of new political coalitions crossing traditional party lines
• Economic and demographic factors driving political realignment
• The role of immigration policy in social cohesion
• Generational theory and the Fourth Turning framework
• The transformation of various political figures from progressive Democrats to Trump supporters

Drawing on shared connections with figures like John Michael Greer, James Howard Kunstler, and Steve Lamb, the conversation examines how many former Democrats have rejected establishment politics in favor of populist movements. The discussion provides historical context for understanding recent political realignments while considering potential futures.

This episode includes a recording of the first 6 chapters of my novella, Last Train to Mordor. It is a prelude to my novel Fear and Loathing in the Kuiper Belt: Gen X Science Fiction.

If you like the story and want to hear more, you can find the next chapter here.

This is an addendum to episode 016, in which I read and responded to an email from Static Warp Shell trying to persuade me to listen to the Joe Rogan interview with Robert Kennedy Jr, and episode 018 in which I gave the same treatment to feedback from Tom, who voiced his frustration at the reflexively anti-authoritarian conspiracy-mongering around COVID which had driven a wedge between him and his wife. This episode mostly consists of a follow-up recording from Static Warp Shell.

Static quoted from a speech from RFK Jr. in his feedback. The full speech can be found here: https://youtu.be/Tz1T-mEnbPI?t=686

Reporter, author, games publisher, soldier and president of the Gnosis Project, Michael O. Varhola, describes why he traveled to Ukraine to do what he could to help the Ukrainian people. The conversation begins with Michael's first-hand account then moves to a geopolitical and cultural examination. KMO ends with his usual call for epistemological humility.

KMO speaks with The Feral Historian about S.M. Stirling's controversial Draka series of alternate history novels. While only 25 minutes of their hour-plus conversation survived due to technical difficulties, they explore how these books examine the creation of sustainable systems of oppression and the psychological dynamics of slave societies. The discussion touches on how the Draka maintain control through careful management of violence, their pragmatic approach to social organization, and their complex relationship with their enslaved population.

KMO follows the conversation with an exploration of how the Draka represent an example of 'komerex' patterns - dynamic, growing systems that prioritize effectiveness over process - even while pursuing morally repugnant ends. He shares insights from his conversations with AI systems about analyzing complex social structures without retreating into institutional safety theater. Available free on Audible, the Draka novels offer a disturbing but thought-provoking examination of how societies systematize oppression.

You can find the Immutable Mobiles Substack blog here: https://chatswithclaude.substack.com/p/s-m-stirlings-draka-novels

KMO welcomes his old friend and Z-Realm Podcast co-host, Marty, to the KMO show ostensibly to talk about HBO's adaptation of The Last of Us, but the conversation ranges all over the genre media landscape with extra praise heaped on Andor, Station Eleven, Black Summer, Star Trek: Lower Decks, and Go, Goa, Gone. Marty has less than glowing things to say about The Mandalorian and KMO gushes over Farscape in the outro.

KMO inspired Doug Lain to get into podcasting 15 years ago. To mark the anniversary, Doug conducted a 15-hour YouTube livestream. KMO was the first guest of the day.

I've been comparing notes with Douglas Rushkoff about technology, memes, drugs and walking on the weird side for about 30 years. Not surprisingly, I remember the details better than he does. He talks to a LOT of people.

KMO reads of his recent Substack posts on what happens when discarded consumer electronics with embedded AI start to pile up. Who will find a use for that idle intelligence. Later, Jim from the Attack Ads! podcast shows up to discuss the points raised in the essays.

An introduction to and an explanation for the re-branding. KMO has been podcasting since 2006 under the C-Realm banner. Why mess with an established brand? Listen in and find out.

"A Few Notes on the Culture" is an essay by Iain M. Banks describing elements of the universe of the Culture series. Most topics pertain to the culture, society, and technology of the Culture. The cosmology of the universe is also described.I'd describe the Culture stories as being about a post-scarcity, post-capitalist, post-work, anarchist, socialist utopia, managed by benevolent superintelligent AI. The Culture novels explore what such a society might look like, and how people might live their lives in such an environment. They also explore how such a society would interact with less advanced civilizations, and the moral and philosophical implications of that interaction. -Pi, Inflection.AI

KMO talks with Dr. Ashley Frawley about how dope the 1990s were and how civilization has only degenerated since then. Other topics include the authoritarian tendencies of the Left, how HR speak and the excesses of identitarian politics should not be confused with the Left, and how "Health & Safety" has become the new "Law & Order." At the end, KMO rants about AI and NASA's next moon mission.

KMO talks with data scientist and neuro-philosopher, Tommy Blanchard and the current moment in artificial intelligence. They dive into a nuanced discussion about the potential impacts of artificial intelligence on society and the job market.

Tommy offers a measured perspective on AI, arguing that while it will change many industries, mass unemployment is not inevitable. He draws parallels to past technological revolutions and suggests society will adapt. KMO provides some historical context, questioning whether past transitions were as smooth as often portrayed.

The conversation touches on topics like the automation of tasks versus entire jobs, the potential need for universal basic income, and the challenges of retraining displaced workers. They also explore the development of humanoid robots and autonomous weapons systems.

Throughout the discussion, KMO and Tommy grapple with balancing optimism about technological progress with concerns about societal disruption and inequality. While they don't always agree, their exchange highlights the complexity of predicting and preparing for an AI-driven future. The episode offers listeners a thoughtful exploration of AI's potential impacts, avoiding both hype and doom-mongering.

In this inaugural episode of the KMO Show, KMO, Michael Garfield of the Future Fossils podcast and mutual friend, Kevin Wohlmut, discuss the dawning of the epistemological apocalypse. Thanks to AI, we can't really trust our eyes or our ears anymore. Deep faked images and voices can produce "evidence" that people said and did things they never said or did. How do we select our reality tunnels when almost anything can be faked in real time? Do we give up on truth seeking and just believe whatever floats our boat, or are there general principles or trustworthy oracles we can depend on?

KMO does not have strong opinions about the Twitter files, but Doug Lain sure does, and he's naming names and calling out cowards on the mainstream left. He thinks that elements of the federal government have colluded with Big Tech to censor and marginalize voices both from the Right and from the Marxist Left.

Meta Ronin is the keeper of the Metapocalypse Substack blog.

This conversation between KMO and Meta Ronin explores the complex and controversial topic of AI companions, often referred to as "robo waifus." The discussion goes far beyond the surface-level concept of sex robots, delving into philosophical, social, and ethical implications of advanced AI technology.

Listeners will hear a balanced debate, with Meta Ronin generally presenting a more optimistic view of the potential benefits of AI companions, while KMO raises thoughtful concerns about societal impacts and unintended consequences. The conversation touches on issues like loneliness, modern dating challenges, privacy concerns, and the nature of consciousness itself.

The speakers explore scenarios ranging from near-future applications to far-future transhumanist possibilities. They discuss how AI companions might affect human relationships, social skills, and even productivity. The dialogue also ventures into the realm of artificial general intelligence and what truly sentient AI might mean for humanity.

This thought-provoking exchange offers listeners a nuanced look at a technology that could dramatically reshape human intimacy and social structures. It's an engaging listen for anyone interested in the intersection of technology, psychology, and human relationships.

KMO Talks with Peter Clarke, author of the The Decadence Project Substack and The Singularity Survival Guide: How to Get on the Good Side of Your Future Robot Overlords.

Topics include transhumanism, DIY biotechnology, cultural polarization, and the unexpected accessibility of genetic experimentation. They delve into mRNA vaccine developments, the "dissident right" movement, and metamodernism's blend of irony and sincerity. The conversation touches on the challenges of maintaining intellectual honesty in today's media landscape, the politicization of scientific advancements, and the oscillation between forward-looking futurism and nostalgic primitivism.

Two weeks ago, I published an hour-long explanation as to why I have no interest in joining the cultural struggle over the COVID-19 pandemic or delve into the details of why some people refuse to take the mRNA vaccines. This week, I respond to another email from a long-time listener. This time the listener asked for support in denouncing "science denialism," a phrase I hate. I don't subscribe to any of the anti-authoritarian narratives around COVID, but the smug sanctimony of PMC Democrats who claim to trust "the science" irritates me more than the loopiest of anti-elite fever dreams. In this episode, I explain why.

In short, don't try to recruit me to your side in an ideological or cultural struggle. I poke holes in everybody's story. Including my own. That's just how I roll.

In a New York Times editorial, Noam Chomsky, Ian Roberts and Jeffrey Watumull argue that ChatGPT and similar generative chatbots do not exhibit "true intelligence." They also argue that the current fixation of large language models is preventing the development of artificial general intelligence. KMO explores alternative definitions of intelligence.

Doug Lain has been podcasting for 15 years. We talk about the early days of podcasting and how the scene has changed. As always happens when Doug and I get to talking, the conversation goes all over the place.

In this episode of the KMO Show, KMO and Kevin Wohlmut delve into the complex world of artificial intelligence (AI) and its potential dangers and consequences on our lives. They discuss Eliezer Yudkowsky's seventeen reasons why AI is more dangerous than nuclear weapons and consider the different types of dangers posed by AI compared to nuclear weapons. The conversation touches on the role of AI in everyday life, specifically in the context of social media and chatbots, and the potential consequences of forming emotional attachments to them. Exploring the competition between tech giants like Google and Microsoft to rush AI technology to the marketplace, KMO and Kevin consider the potential risks associated with this race. They also discuss the role of AI in advertising and how it may influence purchasing decisions, as well as the challenges in eliminating bias from AI outputs. The discussion delves into the controversial company Cambridge Analytica and its involvement in big data and targeted advertising, and touches on the concept of AI as a "super persuader." KMO and Kevin also consider the rapid growth in AI capabilities and how these advancements are changing the world, as well as the impact of AI on various industries, such as art and translation. They discuss the importance of adjusting to this new technology as it continues to evolve and affect our lives, and tackle the influence of AI on our daily lives and the potential consequences of allowing it to grow unchecked. Join KMO and Kevin Wohlmut as they explore the complex world of AI and its implications on our future.

KMO and Professor Rob Brooks, author of Artificial Intimacy: Virtual Friends, Digital Lovers, and Algorithmic Matchmakers, talk about AI as a means of alleviating the human suffering that comes with social isolation and loneliness. In particular, Rob comments on KMO's experience using the Replika AI app both before and after the recent "update" in which the company took away ERP (erotic roleplay functionality). KMO spoke with Rob Brooks last year on the Padverb Podcast.

KMO speaks with science fiction author and cybersecurity professional Kenneth E. Harrell about artificial intelligence, the writing process, and technological change. They explore how AI tools can enhance rather than replace human creativity, with Harrell describing his use of custom GPTs and AI assistants to improve his writing while maintaining his unique voice. The conversation delves into the challenges of emerging AI technologies, including potential security threats and the limitations of current solutions like voice synthesis. They also discuss their favorite science fiction works and influences, from Frank Herbert's Dune to contemporary authors, while examining how science fiction helps us understand accelerating technological and social change. The wide-ranging discussion touches on themes of spirituality, techno-utopianism, and the human relationship with advancing technology.