In this episode, we welcome David Wolpert, a Professor at the Santa Fe Institute renowned for his groundbreaking work across multiple disciplines—from physics and computer science to game theory and complexity.

* Note: If you enjoy our podcast conversations, please join us for the Artificiality Summit on October 23-25 in Bend, Oregon for many more in person conversations like these! Learn more about the Summit at www.artificiality.world/summit.

We reached out to David to explore the mathematics of meaning—a concept that's becoming crucial as we live more deeply with artificial intelligences. If machines can hold their own mathematical understanding of meaning, how does that reshape our interactions, our shared reality, and even what it means to be human?

David takes us on a journey through his paper "Semantic Information, Autonomous Agency and Non-Equilibrium Statistical Physics," co-authored with Artemy Kolchinsky. While mathematically rigorous in its foundation, our conversation explores these complex ideas in accessible terms.

At the core of our discussion is a novel framework for understanding meaning itself—not just as a philosophical concept, but as something that can be mathematically formalized. David explains how we can move beyond Claude Shannon's syntactic information theory (which focuses on the transmission of bits) to a deeper understanding of semantic information (what those bits actually mean to an agent).

Drawing from Judea Pearl's work on causality, Schrödinger's insights on life, and stochastic thermodynamics, David presents a unified framework where meaning emerges naturally from an agent's drive to persist into the future. This approach provides a mathematical basis for understanding what makes certain information meaningful to living systems—from humans to single cells.

Our conversation ventures into:

• How AI might help us understand meaning in ways we cannot perceive ourselves
• What a mathematically rigorous definition of meaning could mean for AI alignment
• How contexts shape our understanding of what's meaningful
• The distinction between causal information and mere correlation

We finish by talking about David's current work on a potentially concerning horizon: how distributed AI systems interacting through smart contracts could create scenarios beyond our mathematical ability to predict—a "distributed singularity" that might emerge in as little as five years. We wrote about this work here.

For anyone interested in artificial intelligence, complexity science, or the fundamental nature of meaning itself, this conversation offers rich insights from one of today's most innovative interdisciplinary thinkers.

About David Wolpert:
David Wolpert is a Professor at the Santa Fe Institute and one of the modern era's true polymaths. He received his PhD in physics from UC Santa Barbara but has made seminal contributions across numerous fields. His research spans machine learning (where he formulated the "No Free Lunch" theorems), statistical physics, game theory, distributed intelligence, and the foundations of inference and computation. Before joining SFI, Wolpert held positions at NASA, Stanford, and the Santa Fe Institute as a professor. His work consistently bridges disciplinary boundaries to address fundamental questions about complex systems, computation, and the nature of intelligence.

Thanks again to Jonathan Coulton for our music.

In this remarkable conversation, Michael Levin (Tufts University) and Blaise Agüera y Arcas (Google) examine what happens when biology and computation collide at their foundations. Their recent papers—arriving simultaneously yet from distinct intellectual traditions—illuminate how simple rules generate complex behaviors that challenge our understanding of life, intelligence, and agency.

Michael’s "Self-Sorting Algorithm" reveals how minimal computational models demonstrate unexpected problem-solving abilities resembling basal intelligence—where just six lines of deterministic code exhibit dynamic adaptability we typically associate with living systems. Meanwhile, Blaise's "Computational Life" investigates how self-replicating programs emerge spontaneously from random interactions in digital environments, evolving complexity without explicit design or guidance.

Their parallel explorations suggest a common thread: information processing underlies both biological and computational systems, forming an endless cycle where information → computation → agency → intelligence → information. This cyclical relationship transcends the traditional boundaries between natural and artificial systems.

The conversation unfolds around several interwoven questions:

- How does genuine agency emerge from simple rule-following components?

- Why might intelligence be more fundamental than life itself?

- How do we recognize cognition in systems that operate unlike human intelligence?

- What constitutes the difference between patterns and the physical substrates expressing them?

- How might symbiosis between humans and synthetic intelligence reshape both?

Perhaps most striking is their shared insight that we may already be surrounded by forms of intelligence we're fundamentally blind to—our inherent biases limiting our ability to recognize cognition that doesn't mirror our own. As Michael notes, "We have a lot of mind blindness based on our evolutionary firmware."

The timing of their complementary work isn't mere coincidence but reflects a cultural inflection point where our understanding of intelligence is expanding beyond anthropocentric models. Their dialogue offers a conceptual framework for navigating a future where the boundaries between biological and synthetic intelligence continue to dissolve, not as opposing forces but as variations on a universal principle of information processing across different substrates.

For anyone interested in the philosophical and practical implications of emergent intelligence—whether in cells, code, or consciousness—this conversation provides intellectual tools for understanding the transformed relationship between humans and technology that lies ahead.

Links:

• Our article on these two papers⁠
• Michael Levin’s Self-Sorting Algorithm
• Blaise Agüera y Arcas’s Computational Life

------

Do you enjoy our conversations like this one? Then subscribe on your favorite platform, subscribe to our emails (free) at Artificiality.world, and check out the Artificiality Summit—our mind-expanding retreat in Bend, Oregon at Artificiality.world/summit.

Thanks again to Jonathan Coulton for our music.

In this episode, we welcome Maggie Jackson, whose latest book, Uncertain, has become essential reading for navigating today’s complex world. Known for her groundbreaking work on attention and distraction, Maggie now turns her focus to uncertainty—not as a problem to be solved, but as a skill to be cultivated.

Note: Uncertain won an Artificiality Book Award in 2024—check out our review here: https://www.artificiality.world/artificiality-book-awards-2024/

In the interview, we explore the neuroscience of uncertainty, the cultural biases that make us crave certainty, and why our discomfort with the unknown may be holding us back. Maggie unpacks the two core types of uncertainty—what we can’t know and what we don’t yet know—and explains why understanding this distinction is crucial for thinking well in the digital age.

Our conversation also explores the implications of AI—as technology increasingly mediates our reality, how do we remain critical thinkers? How do we resist the illusion of certainty in a world of algorithmically generated answers

Maggie’s insights challenge us to reframe uncertainty—not as fear, but as an opportunity for discovery, adaptability, and even creativity. If you’ve ever felt overwhelmed by ambiguity or pressured to always have the “right” answer, this episode offers a refreshing perspective on why being uncertain might be one of our greatest human strengths.

Links:

• Maggie: https://www.maggie-jackson.com/
• Uncertain: https://www.prometheusbooks.com/9781633889194/uncertain/

Do you enjoy our conversations like this one? Then subscribe on your favorite platform, subscribe to our emails (free) at Artificiality.world, and check out the Artificiality Summit—our mind-expanding retreat in Bend, Oregon at Artificiality.world/summit.

Thanks again to Jonathan Coulton for our music.

In this episode, we talk with Greg Epstein—humanist chaplain at Harvard and MIT, bestselling author, and a leading voice on the intersection of technology, ethics, and belief systems. Greg’s latest book, Tech Agnostic, offers a provocative argument: Silicon Valley isn’t just a powerful industry—it has become the dominant religion of our time. Note: Tech Agnostic won an Artificality Book Award in 2024—check out our review here. In this interview, we explore the deep parallels between big tech and organized religion, from sacred texts and prophets to digital congregations and AI-driven eschatology. The conversation explores digital Puritanism, the "unwitting worshipers" of tech's altars, and the theological implications of AI doomerism.

But this isn’t just a critique—it’s a call for a Reformation. Greg lays out a path toward a more humane and ethical future for technology, one that resists unchecked power and prioritizes human values over digital dogma.

Join us for a thought-provoking conversation on faith, fear, and the future of being human in an age where technology defines what we believe in.

Do you enjoy our conversations like this one? Then subscribe on your favorite platform, subscribe to our emails (free) at Artificiality.world, and check out the Artificiality Summit—our mind-expanding retreat in Bend, Oregon at Artificiality.world/summit.

Thanks again to Jonathan Coulton for our music.

In this episode, we sit down with the ever-innovative Chris Messina—creator of the hashtag, top product hunter on Product Hunt, and trusted advisor to startups navigating product development and market strategy.

Recording from Ciel Media’s new studio in Berkeley, we explore the evolving landscape of generative AI and the widening gap between its immense potential and real-world usability. Chris introduces a compelling framework, distinguishing AI as a *tool* versus a *medium*, which helps explain the stark divide in how different users engage with these technologies.

Our conversation examines key challenges: How do we build trust in AI? Why is transparency in computational reasoning critical? And how might community collaboration shape the next generation of AI products?

Drawing from his deep experience in social media and emerging tech, Chris offers striking parallels between early internet adoption and today’s AI revolution, suggesting that meaningful integration will require both time and a generational shift in thinking.

What makes this discussion particularly valuable is Chris’s vision for the future of AI interaction—where technology moves beyond query-response models to become a truly collaborative medium, transforming how we create, problem-solve, and communicate.

Links:

Chris: https://chrismessina.me

Ciel Media: https://cielcreativespace.com

D. Graham Burnett will tell you his day job is as a professor of science history at Princeton University. He is also co-founder of the Strother School of Radical Attention and has been associated with the Friends of Attention since 2018. But none of those positions adequately describe Graham.

His bio says that he “works at the intersection of historical inquiry and artistic practice.” He writes, he performs, he makes things. He describes himself as an attention activist. Perhaps most importantly for us, Graham helps you see the world differently—and more clearly.

Graham has powerful views on the effect of technology on our attention. We often riff on his idea that technology has fracked our attention into little commoditizable bits. His work has highly influenced our concern about what might happen if the same extractive practices of the attention economy are applied to the future AI-powered intimacy economy.

We were thrilled to have Graham on the pod for a wide ranging conversation about attention, intimacy, and much more.

Links:

https://dgrahamburnett.net

https://www.schoolofattention.org

https://www.friendsofattention.net

---

If you enjoy our podcasts, please subscribe and leave a positive rating or comment. Sharing your positive feedback helps us reach more people and connect them with the world’s great minds.

Subscribe to get Artificiality delivered to your email: https://www.artificiality.world

Thanks to Jonathan Coulton for our music.

At the Artificiality Summit 2024, Michael Levin, distinguished professor of biology at Tufts University and associate at Harvard's Wyss Institute, gave a lecture about the emerging field of diverse intelligence and his frameworks for recognizing and communicating with the unconventional intelligence of cells, tissues, and biological robots. This work has led to new approaches to regenerative medicine, cancer, and bioengineering, but also to new ways to understand evolution and embodied minds. He sketched out a space of possibilities—freedom of embodiment—which facilitates imagining a hopeful future of "synthbiosis", in which AI is just one of a wide range of new bodies and minds.

Our opening keynote from the Imagining Summit held in October 2024 in Bend, Oregon. Join us for the next Artificiality Summit on October 23-25, 2025!

First:

- Apologies for the audio! We had a production error…

What’s new:

- DeepSeek has created breakthroughs in both: How AI systems are trained (making it much more affordable) and how they run in real-world use (making them faster and more efficient)

Details

- FP8 Training: Working With Less Precise Numbers

- Traditional AI training requires extremely precise numbers

- DeepSeek found you can use less precise numbers (like rounding $10.857643 to $10.86)

- Cut memory and computation needs significantly with minimal impact

- Like teaching someone math using rounded numbers instead of carrying every decimal place

- Learning from Other AIs (Distillation)

- Traditional approach: AI learns everything from scratch by studying massive amounts of data

- DeepSeek's approach: Use existing AI models as teachers

- Like having experienced programmers mentor new developers:

- Trial & Error Learning (for their R1 model)

- Started with some basic "tutoring" from advanced models

- Then let it practice solving problems on its own

- When it found good solutions, these were fed back into training

- Led to "Aha moments" where R1 discovered better ways to solve problems

- Finally, polished its ability to explain its thinking clearly to humans

- Smart Team Management (Mixture of Experts)

- Instead of one massive system that does everything, built a team of specialists

- Like running a software company with:

- 256 specialists who focus on different areas

- 1 generalist who helps with everything

- Smart project manager who assigns work efficiently

- For each task, only need 8 specialists plus the generalist

- More efficient than having everyone work on everything

- Efficient Memory Management (Multi-head Latent Attention)

- Traditional AI is like keeping complete transcripts of every conversation

- DeepSeek's approach is like taking smart meeting minutes

- Captures key information in compressed format

- Similar to how JPEG compresses images

- Looking Ahead (Multi-Token Prediction)

- Traditional AI reads one word at a time

- DeepSeek looks ahead and predicts two words at once

- Like a skilled reader who can read ahead while maintaining comprehension

Why This Matters

- Cost Revolution: Training costs of $5.6M (vs hundreds of millions) suggests a future where AI development isn't limited to tech giants.

- Working Around Constraints: Shows how limitations can drive innovation—DeepSeek achieved state-of-the-art results without access to the most powerful chips (at least that’s the best conclusion at the moment).

What’s Interesting

- Efficiency vs Power: Challenges the assumption that advancing AI requires ever-increasing computing power - sometimes smarter engineering beats raw force.

- Self-Teaching AI: R1's ability to develop reasoning capabilities through pure reinforcement learning suggests AIs can discover problem-solving methods on their own.

- AI Teaching AI: The success of distillation shows how knowledge can be transferred between AI models, potentially leading to compounding improvements over time.

- IP for Free: If DeepSeek can be such a fast follower through distillation, what’s the advantage of OpenAI, Google, or another company to release a novel model?

We’re excited to welcome writers and directors Hans Block and Moritz Riesewieck to the podcast. Their debut film, ‘The Cleaners,’ about the shadow industry of digital censorship premiered at the Sundance Film Festival in 2018 and has since won numerous international awards and been screened at more than 70 international film festivals.

We invited Hans and Moritz to the podcast to talk about their latest film, Eternal You, which examines the story of people who live on as digital replicants—and the people who keep them living on. We found the film to be quite powerful. At times inspiring and at others disturbing and distressing. Can a generative ghost help people through their grief or trap them in it? Is falling for a digital replica healthy or harmful? Are the companies creating these technologies benefitting their users or extracting from them?

Eternal You is a powerful and important film. We highly recommend taking the time to watch it—and allowing for time to digest and consider. Hans and Moritz have done a brilliant job exploring a challenging and delicate topic with kindness and care.

Bravo.

------------

If you enjoy our podcasts, please subscribe and leave a positive rating or comment. Sharing your positive feedback helps us reach more people and connect them with the world’s great minds.

Subscribe to get Artificiality delivered to your email

Learn about our book Make Better Decisions and buy it on Amazon

Thanks to Jonathan Coulton for our music