Who writes the next space treaty? We put the question to Harvard economist Matthew Weinzierl and Blue Origin's Brendan Rosseau, co-authors of Space to Grow. Neither wants a master plan.

Weinzierl would start by gathering 25 to 30 leading thinkers from start-ups, government and non-profits. Yet he is sceptical that any room can design the future of space from the top down. He prefers order that emerges on its own.

Rosseau begins with the end state: competition that drives progress, no military conflict, and rules of the road like those that govern the sea. He doubts one grand treaty could deliver that, and suggests a little strategic ambiguity may even help.

Please enjoy the show. And subscribe if you do.

Full Episode & Links
• Watch on YouTube: https://www.youtube.com/watch?v=tnInzxdiFkk&t=1363s

Connect with Thinking on Paper
🌐 Website: https://www.thinkingonpaper.xyz
📷 Instagram: @toptechpodcast
🎧 Listen on Spotify: Listen on Spotify
🍎 Listen on Apple Podcasts: Listen on Apple Podcasts

Chapters

(00:00) You're in charge of space. Now what?
(01:04) Matt: "I like emergent order"
(02:39) Brendan: competition yes, war no
(04:41) Why not one grand treaty?
(05:22) The four-word rule for space

#SpaceTreaty #SpaceToGrow #ThinkingOnPaper

"AI is a very poor writer," says celebrated author, hip-hop historian, and Bruce Lee biographer Jeff Chang.

While large language models can mirror sentence structures or copy paragraphs, Chang argues that the raw output is garbage.

He also shares insights from his interview with legendary hip-hop producer DJ Shadow on his KALW radio show, Notes from the Edge.

Shadow calls generative AI music "corny" and warns that those hyping artificial intelligence tools are usually the ones profiting from them.

This leads to the ultimate question for modern creators: If you rely on AI for ideation, are you actually being creative?

In this episode, we cover:

• Why Jeff Chang writes every word himself and limits AI to research
• The craft of revision: Why great non-fiction requires four drafts
• Insights from Karen Hao, DJ Shadow, and the commercial hype around AI tools
• How analog constraints, turntables, and four-track recording built modern sampling culture
• The difference between machine-generated ideas and human artistic judgment
• Could an 18-year-old produce the next Endtroducing..... using generative AI models?

Please enjoy the show.

Connect with Thinking on Paper

🌐 Website & Show Notes: thinkingonpaper.xyz
📷 Instagram: @toptechpodcast
🎧 Spotify: Listen on Spotify
🍎 Apple Podcasts: Listen on Apple Podcasts

Chapters

(00:00) Does Jeff Chang Use AI to Write?
(01:33) The Revision Process: Why Great Writing Takes Four Drafts
(04:51) Karen Hao, DJ Shadow, and Why AI Music Feels "Corny"
(07:18) What Analog Sampling Culture Teaches Us About AI
(11:32) Ideation vs. Artistry: Can You Be Creative If AI Does the Thinking?
(13:20) Could an 18-Year-Old Make the Next Endtroducing..... with Generative AI?

..

#jeffchang #djshadow #generativeai #hiphopculture #sampling #writingcraft #creativity #thinkingonpaper

If you want superintelligence inside robots and autonomous agents, a Large Language Model (LLM) isn't enough, AI needs a true world model.

In this episode of Thinking On Paper, Mark and Jeremy get their heads around world models, how they work, and why they represent the next frontier of superintelligence.

Pim de Witte, CEO of General Intuition, recently sat down with Meta Chief AI Scientist Yann LeCun at Station F in Paris to frame the limits of current AI.

De Witte compared LLMs to a snowball rolling downhill: it can't see the crash coming because "its entire world is just itself."

Meanwhile, traditional video generation models act like a dream, you watch the frames unfold, but you can't interact or change the outcome. A true world model acts like a lucid dream: an interactive simulation where the agent acts, and the model accurately predicts what happens next.

Jeremy also explains why game studios still struggle to code believable stadium crowd physics fan-by-fan, and how spatial-temporal world models can learn emergent crowd dynamics instead.

We discuss:

• Video Models vs. World Models: Why visual generation is passive "dreaming" while world modeling is interactive "lucid dreaming"
• The Snowball Metaphor: Why LLMs lack true physical perception and spatial reasoning
• Epistemic Uncertainty: How AI handles what it hasn't seen in its training data
• Aleatoric Uncertainty: Managing real-world randomness that can never be eliminated
• Diffusion vs. V-JEPA: Comparing image-based diffusion predictors to Yann LeCun's Joint-Embedding Predictive Architecture
• The Backrooms Effect: Why AI video models generate blurry faces and show visual doubt when predicting futures

Please enjoy the show.

Connect with Thinking on Paper

🌐 Website & Show Notes: thinkingonpaper.xyz
📷 Instagram: @toptechpodcast
🎧 Spotify: Listen on Spotify
🍎 Apple Podcasts: Listen on Apple Podcasts

Chapters

(00:00) Video Models vs. World Models: Dreaming vs. Lucid Dreaming
(00:41) Why Video Game Stadium Crowds Still Look Fake
(01:44) Can Compute Alone Recreate Realistic Stadium Crowds?
(02:09) Predicting Multi-Agent Outcomes with Spatial AI
(02:55) LLMs as Snowballs: The Lack of Physical Perception
(04:00) Epistemic Uncertainty: What the Model Hasn't Seen
(04:35) Aleatoric Uncertainty: Managing Inherent Randomness
(05:16) The Backrooms & Blurry Visuals: How Models Express Doubt
(06:24) Diffusion vs. JEPA: Two Architectures for Predicting the Future
(06:55) How Generative Image Models Turn Noise into Reality

#worldmodels #artificialintelligence #yannlecun #pimdewitte #robotics #generalintuition #spatialai #jepa #thinkingonpaper