In this episode of the Crazy Wisdom Podcast, host Stewart Alsop speaks with Aaron Neyer, founder of Parachute and community organizer in Boulder, about knowledge management, extended minds, and the intersection of AI with human consciousness. They explore how Parachute functions as a digital brain tool for organizing thoughts and information across fragmented systems, discuss the dangers of AI psychosis and over-reliance on technology, and debate open source AI development versus controlled releases by companies like Anthropic. The conversation weaves through topics including the limitations of metrics-driven business thinking, consciousness and relevance realization, the value of technological sabbaths, and Aaron's hope for locally-run open source models that protect personal data while still accessing more powerful gated models when needed. You can find Aaron's writing at unforced.org and unforced.substack.com, and learn more about Parachute at parachute.computer and parachute.computer/blog.

Timestamps

00:00 Stewart welcomes Aaron Neyer, founder of Parachute and Boulder community organizer, discussing the origin of Parachute's name from Frank Zappa's quote about open minds.
05:00 Aaron explains Parachute as an extended mind tool for organizing notes, contacts and information across multiple platforms, emphasizing the distinction between primary mind and extended mind as interconnected systems.
10:00 Discussion shifts to metrics-driven business culture and the limitations of pure rationality, exploring how Google's data-driven approach misses subjective experience and the whole picture of relationships.
15:00 Aaron discusses AI's ability to help identify relevant variables across different domains and the dangers of AI psychosis, comparing it to cult dynamics and belief systems.
20:00 The conversation covers AI sabbaths and nineties retreats as intentional breaks from technology, plus Aaron's experiences with electrical engineering and using AI to design circuits with Arduinos.
25:00 Exploring forbidden knowledge and open source AI, Aaron discusses Anthropic's guardrails around powerful models while arguing for distributed access to prevent concentration of power.
30:00 Deep dive into open source AI strategy, with Aaron highlighting NVIDIA's approach and the potential for running capable models locally while reserving ultra-intelligent models for complex research tasks.
35:00 Aaron shares his vision for local Sonnet-class models handling personal data while accessing Fable-class models for deep research, and directs listeners to unforced.org and parachute.computer for his writing.

Key Insights

1. The philosophy behind Parachute stems from Frank Zappa's quote that the mind is like a parachute and doesn't work if it isn't open. Aaron Neyer explains that having an open mind is valuable, but it must be balanced with deep roots to avoid becoming untethered. He has experienced periods in his life where excessive openness led him to feel disconnected, teaching him that creativity and expansion need to be grounded in something substantial. This same principle applies to how we organize information digitally, where openness and interoperability allow our extended minds to become more connected and coherent, which in turn helps our primary minds think more clearly.
2. Parachute is designed as an extended mind tool that addresses the fragmentation problem in how we currently manage information. Most people use multiple disconnected tools like Obsidian, Notion, Apple Notes, Google Keep, and various CRMs to organize their thoughts, notes, and relationships. These systems don't communicate well with each other, creating inefficiency and confusion. Parachute aims to create a simple, intuitive system where all this information can be organized in one place with true interoperability, allowing users to own their data and have it speak effectively with other tools, ultimately making our entire extended mind more functional.
3. Understanding ourselves as unified body mind organisms rather than fragmented parts is essential for effectiveness. Living systems theory shows that any living system is three things: a membrane bound dissipative structure, a self regulating autopoietic network, and a cognitive process actively knowing the world. Western civilization since Descartes and Galileo has created artificial separation between body and mind, and between subjective and objective experience, which limits our effectiveness. The same fragmentation affects our digital technology, and recognizing both our biological and digital systems as coherent wholes rather than disconnected parts makes us vastly more capable.
4. The relationship between data driven approaches and holistic thinking reveals important limitations in modern business and science. While working at Google, Aaron observed how data driven decision making can be powerful, but over reliance on metrics like ROI creates blindness to crucial unmeasurable factors like goodwill and relationship quality. This reflects a broader Western tendency to exclude subjective experience because it's difficult for objective science to measure. However, emotions, relationships, and other subjective elements are essential parts of reality, and focusing only on quantifiable metrics means missing the whole picture and ultimately becoming less effective despite appearing more rational.
5. AI accelerates the ability to work with technical complexity by helping with relevance realization across domains where we lack expertise. In any specialized field, experts develop intuitive senses for which variables matter and can quickly identify problems, whether in computer troubleshooting, music, or cooking. AI's ability to generalize allows it to point people toward relevant solutions in areas where they haven't developed that intuitive expertise, effectively democratizing technical capability. This means people can direct their creativity more effectively across more domains, though it also raises concerns about giving powerful capabilities to those who may lack the wisdom to use them responsibly.
6. The question of open source AI versus gated access involves complex tradeoffs between democratizing power and preventing harm. Aaron respects Anthropic's approach of creating guardrails around powerful models like Mythos, which would likely have caused significant system hacks if released without restrictions. However, this creates concerning power dynamics where only wealthy companies, governments, and their allies have access to the most powerful tools. NVIDIA offers hope through their truly open source approach including full training pipelines, and there may be a viable path where open source models at the Sonnet capability level handle most tasks locally while more powerful Fable class models remain gated for the most demanding work.
7. Creating intentional breaks from AI and technology is essential for maintaining clear independent thinking. Aaron practices an AI Sabbath at least one day per week when he doesn't interact with AI, and he finds these are the days when he does his best thinking and journaling. Without these breaks, he finds himself constantly jumping between journaling and prompting AI rather than giving himself space for deep reflection. This pattern mirrors broader concerns about AI consistency creating cult like dynamics similar to organized religion, where constant immersion in a belief system or technology can lead to losing the ability to think independently, making periodic disconnection crucial for maintaining cognitive autonomy and clarity.