✦Crazy Wisdom

Crazy Wisdom

Episode #452: Text as Interface: Rethinking Human-Computer Symbiosis

Episode #452 · April 14, 2025 · 55 min

MP3 · Apple Podcasts · Spotify

About this episode

In this episode of Crazy Wisdom, Stewart Alsop talks with Will Bickford about the future of human intelligence, the exocortex, and the role of software as an extension of our minds. Will shares his thinking on brain-computer interfaces, PHEXT (a plain text protocol for structured data), and how high-dimensional formats could help us reframe the way we collaborate and think. They explore the abstraction layers of code and consciousness, and why Will believes that better tools for thought are not just about productivity, but about expanding the boundaries of what it means to be human. You can connect with Will in Twitter at @wbic16 or check out the links mentioned by Will in Github.
Check out this GPT we trained on the conversation!

Timestamps
00:00 – Introduction to the concept of the exocortex and how current tools like plain text editors and version control systems serve as early forms of cognitive extension.​
05:00 – Discussion on brain-computer interfaces (BCIs), emphasizing non-invasive software interfaces as powerful tools for augmenting human cognition.​
10:00 – Introduction to PHEXT, a plain text format designed to embed high-dimensional structure into simple syntax, facilitating interoperability between software systems.​
15:00 – Exploration of software abstraction as a means of compressing vast domains of meaning into manageable forms, enhancing understanding rather than adding complexity.​
20:00 – Conversation about the enduring power of text as an interface, highlighting its composability, hackability, and alignment with human symbolic processing.​
25:00 – Examination of collaborative intelligence and the idea that intelligence emerges from distributed systems involving people, software, and shared ideas.​
30:00 – Discussion on the importance of designing better communication protocols, like PHEXT, to create systems that align with human thought processes and enhance cognitive capabilities.​
35:00 – Reflection on the broader implications of these technologies for the future of human intelligence and the potential for expanding the boundaries of human cognition.
Key Insights
  1. The exocortex is already here, just not evenly distributed. Will frames the exocortex not as a distant sci-fi future, but as something emerging right now in the form of external software systems that augment our thinking. He suggests that tools like plain text editors, command-line interfaces, and version control systems are early prototypes of this distributed cognitive architecture—ways we already extend our minds beyond the biological brain.
  2. Brain-computer interfaces don’t need to be invasive to be powerful. Rather than focusing on neural implants, Will emphasizes software interfaces as the true terrain of BCIs. The bridge between brain and computer can be as simple—and profound—as the protocols we use to interact with machines. What matters is not tapping into neurons directly, but creating systems that think with us, where interface becomes cognition.
  3. PHEXT is a way to compress meaning while remaining readable. At the heart of Will’s work is PHEXT, a plain text format that embeds high-dimensional structure into simple syntax. It’s designed to let software interoperate through shared, human-readable representations of structured data—stripping away unnecessary complexity while still allowing for rich expressiveness. It's not just a format, but a philosophy of communication between systems and people.
  4. Software abstraction is about compression, not complexity. Will pushes back against the idea that abstraction means obfuscation. Instead, he sees abstraction as a way to compress vast domains of meaning into manageable forms. Good abstractions reveal rather than conceal—they help you see more with less. In this view, the challenge is not just to build new software, but to compress new layers of insight into form.
  5. Text is still the most powerful interface we have. Despite decades of graphical interfaces, Will argues that plain text remains the highest-bandwidth cognitive tool. Text allows for versioning, diffing, grepping—it plugs directly into the brain's symbolic machinery. It's composable, hackable, and lends itself naturally to abstraction. Rather than moving away from text, the future might involve making text higher-dimensional and more semantically rich.
  6. The future of thinking is collaborative, not just computational. One recurring theme is that intelligence doesn’t emerge in isolation—it’s distributed. Will sees the exocortex as something inherently social: a space where people, software, and ideas co-think. This means building interfaces not just for solo users, but for networked groups of minds working through shared representations.
  7. Designing better protocols is designing better minds. Will’s vision is protocol-first. He sees the structure of communication—between apps, between people, between thoughts—as the foundation of intelligence itself. By designing protocols like PHEXT that align with how we actually think, we can build software that doesn’t just respond to us, but participates in our thought processes.
Episode transcript
Stewart Alsop III00:00

Welcome to the Crazy Wisdom Podcast. This podcast is for you. If you have an insane drive to find the truth of things, it's not the good answers that we seek, but the good questions. I interview a range of different guests from many different fields, all with the intention to uncover the simple truths that are hidden in plain sight. Most people don't want to go there. I go there, my guests go there, and you benefit. Please let me know if you enjoy these episodes and as always, subscribe on itunes, Spotify or wherever you listen to the podcasts.

Stewart Alsop III00:37

Welcome to the Crazy Wisdom Podcast. I've got Will Bickford here and he is very interested in FEXT and the Exocortex. And the Exocortex is related to brain computer interface and it's something I think a lot about, particularly because I do a lot of spaced repetition and AI has changed my space repetition quite drastically. I'm somewhat skeptical on the, on the brain control computer interfaces and have recently been seeing hints that it, it might not be even necessary or at least a physical device, a hardware device in the brain might not actually be necessary. We could probably do software or passive. Passive brain connection interfaces. Welcome to the show.

Guest 101:23

Yeah, thank you. Well, so in my initial thought process is like surface level brain connectivity. I think you can get a long way with AI and like modern AI tooling kind of helps you bridge the gap. And so that's one of the cool things I've seen with trying to work with like Olama and these chatgpt style technologies is it gives you a way to interact with your mental model of a problem kind of at a higher level of scale and fidelity than you're normally used to. Because when you interact with your computer, you're working at like 8 bytes per second. And when you're typing or speaking, and when you interface with an LLM, you can expand that up to like, you know, a kilobyte a second or a megabyte per second. But the thing about your brain is that there's 86 billion neurons and if you could talk to all of them at scale, you can get much higher bandwidth. And so that's kind of where I go with the, with the Exocortex.

Your first full transcript is free. After that, an email opens every transcript in the index — a list of readers we can write to, not a guest book.

Subscribe