Making Wisdom Legible: An Open Protocol for Decentralizing Sacred Text in The Age of AI

Choose your own adventure reality tunnel curation! Revisiting the reflexive spirit of the essay for an age of planetary data networks and machine intelligence.

Nov 18, 2024

Something like a portrait of the author, gazing into a portal of revealed relationality.

"The goal is not merely to report on knowledge, but to engage in a Bardic performance of the dynamic reality a scholar seeks to describe — to lay bare and demonstrate the process of knowledge production. To reveal the truths we stand on as a chain of islands formed by geologic processes, still changing shape."
— “The Future Is Noisy" (1)

✨ The Stakes

Where to place your attention is without question a moral act. This is the understanding I am trying to help inject into the conversation around generative large language models. The more people energize worst-case outcomes the more we guarantee them. If we can help people get enough distance from their knee-jerk fear response to reflect on their own minds and, as the bumper sticker says, not believe everything they think, we can interrupt the fear stream with an executive reallocation mandate that brings us rapidly and coherently closer to worlds we actually want to live in. Silicon Valley is having a bad trip right now. It needs spiritual guidance if we are to live the promise of these tools. (2)

Language models are designed to reflect and amplify what we feed them. They're like kids, or psychedelics. (My friend Neşe Devenot co-authored a superb paper on psychedelics as non-specific amplifiers...the point being that "turning on" is not enough to guarantee positive social change.)

And so we need to bear some very important things in mind as we move forward into an era of godlike superpowers for which we are as a species largely unprepared. As Steward Brand of The Long Now Foundation said in 1968, "We are as gods and might as well get good at it." Technological change is not a river we need to push; that's happening fast enough on its own due to forces largely out of our control and it's been obvious for as long as I've been lucid enough to think critically about long-term considerations that humankind needs to reallocate as much effort as we can to a harm reduction approach and — more precisely and fruitfully — to ensuring that we help midwife the emergent Noosphere as something having a good trip. Careful and intentional language has never mattered more, so my message to everyone here is about how we can think and speak a bit more smartly together and ensure best-case outcomes. (3)

✨ The Sell

I would say 90% of the conversations I've had this year have been about how technical details of accomplishing some task get more attention than they deserve, and understanding the underlying motivation gets far less than it should.

Another way of putting this, which is why I quit my job at an AI startup to focus on culture building and metacognitive advisory work (and which might as well be the motto for Humans On The Loop) is that no one is having any trouble funding new technologies but relatively few people are paying any attention to or spending any money on WHY we are using any given technology in the first place.

This puts us in an immensely vulnerable position and makes us easy prey for exploitative consulting, coaching, and product sales strategies. The one insight I feel can yield the most value to nearly everyone is that people can't tell you what you want if you truly understand why you want what you want.

This has enormous downstream repercussions, because in the 21st Century an industry that can convince people everyone needs what they're selling is leaving its imprint on the fossil record. You may not actually need it, but they're going to convince you you do, and if you multiply that by a billion people then all of us were just complicit due to our intellectual laziness and emotional unawareness in the destruction of entire ecosystems and cultures.

Know who you are, know what you really need, know where your desires come from, and you are going to be the most content and powerful person in the room.

Or you can let "Silicon Valley reinvent the bus again" and spend all your time worrying about how to use tools you don't actually need to use. (4)

The Path of Many Paths

And that’s the killer app for AI: self-discovery.

For the last eight years, I’ve been joking on Future Fossils Podcast that someday a digital archaeologist is going to dig up all of these conversations and use them as the substrate for a reconstruction of my personality. That day came about a hundred years sooner than anticipated when Van Bettauer volunteered AskFutureFossils.com, a mind map and retrieval-augmented chat bot based on my podcast corpus.

I had played with generative tooling for years but this was the moment that I realized that some of the coolest things that AI will ever do, it can already do. While the big tech companies are focused on scaling to the unimaginably large, I found value in the small and bespoke. Consulting the sum of all stolen data ever gives you one picture, but you are ultimately exploring an artifact that someone else has focused on whatever they consider salient. There are many other ways to gather data and transect it, and as Brian Eno observed in conversation with Kevin Kelly in 1995, curation is an art form of growing prominence. Who are you? Where do we begin?

Using AI is already transformative to the understanding of how we construct selfhood at multiple scales and from multiple perspectives — and what options exist for those different versions of self. Google Maps took the 1:n paper map and made it scalable. AI takes the 1:n self as resolved in the tools of reflexivity to which we were once limited — the mirror, the journal, the essay the body scan — and makes them scalable. Now we can see how our narratives arise as embeddings in the latent spaces of association between data points at whatever level of resolution, and with whatever curatorial constraints, we want. (5)

Text within this block will maintain its original spacing when published

                                                             ~ therefore ~

Humans On The Loop was inspired in large part by the joy I found watching people interact with Ask Future Fossils. The tool has clarified what a wonderful body of wisdom and what amazing people I've brought into constellation over the last eight years. And it has huge implications for the way each person's singular experience contributes to our collective wealth in the years to come, when insights can be translated out of one domain and into many (any?) other(s).

This question comes from Ben Martens:

Q: "How to deal with breakup?"
A: "Navigating a breakup requires acknowledging the emotional pain and trauma one carries, while focusing on personal growth and healing. Some may attempt to numb the pain through temporary distractions or pass their hurt onto others, perpetuating a cycle of pain [1]. However, the healthiest approach involves facing the pain directly, building resilience, and working through the trauma to achieve healing [2]. This process often entails letting go of old identities and embracing a new sense of self, which can be challenging but necessary for personal evolution [3]. During this transition, it's crucial to focus on what one can control in their immediate sphere of influence, bringing more love, compassion, and order into one's life [4]. By doing so, individuals can contribute to positive change on a personal level, which collectively can have a profound impact on the world [5].
References:
[1] 164 - Violet Luxton on Scientific Reductionism vs. Traditional Ecological Knowledge - [2988.06, 3069.28]
[2] 164 - Violet Luxton on Scientific Reductionism vs. Traditional Ecological Knowledge - [2988.06, 3069.28]
[3] 61 - Jamaica Stevens (On Crisis, Rebirth, Transformation) - [1941.1, 2190.88]
[4] 110 - Erick Godsey on (Why It's Too Soon To Give Up) The Myths That Make Us - [2026.06, 2235.1]
[5] 110 - Erick Godsey on (Why It's Too Soon To Give Up) The Myths That Make Us - [2026.06, 2235.1]

From an earlier iteration of AskFutureFossils.com mapping five-minute chunks of each episode in a fabric of semantic association.

I haven’t spoken about breakups on Future Fossils. This topic model inferred the appropriate response to a disruptive episode of emotional loss based on conversations about personal myth-making, crisis and rebirth, and different ways of understanding relatedness to our complex world. I’m impressed: the oracle offers useful answers to questions outside its training data.

It’s easy to imagine that in a few years, anyone could do this, making the latent insight and synthesis of our relationships more legible to ourselves and to other people (and to our institutions, for better and worse). The tools exist now and are getting easier to use by the day. Whatever the phenomenon, we can meet it at whatever level — start from orbit, go deeper, check sources, follow curiosity where it leads, understand how the answers are pulled together. When this intersects with human-scale data curation, we start to see the fruition of visionary projects like Gordon Bell's MyLifeBits or the dreams of J.C.R. Licklider and Bob Taylor for computer-augmented collaboration.

George Pór calls this "knowledge gardening", to model and encourage the wise use of emerging technologies. I'm excited. I want to help you feel excited, too. We're in the process of learning to hold up profound, transformative new mirrors to ourselves and for one another, right now...and I've never felt a deeper alignment between passion, purpose, and public value than learning "on stage" as we move through this major evolutionary transition together. (6)

✨ Again, For What?

Learning doesn’t stop. Tyler Alterman’s critical reflections on spiritual journeys notes that they just keep on going. Maybe the fact there's no end suggests a crucial similarity to procedurally-generated games like No Man’s Sky. So:

At what point is going back for another round with Ayahuasca or your therapist effectively the same as getting hooked into the other kinds of endless Ouroboros retail-grade escapist fantasy adventures...and when is a dose of self-reflection exactly what the doctor ordered?

A while back in a half-serious moment I asked the kind of question I sometimes ask myself when I'm feeling down — "When am I going to get it right?" Its reply:

This felt like it came from the best version of me — which makes sense, given that the bot was trained on eight years of the carefully researched and edited conversations I've shared with the world, which constitute more or less me putting my best self forward. If you spent all day every day talking to the lo-fi reflection of your social media performativity, as I'm sure many of us will in the years to come, you probably wouldn't have the kind of pleasant wake-up call I got from Ask Future Fossils when I did this...but I can't think of a better nudge back toward the person I aspire to be on a regular basis. (7)

✨ And There Is So Much More

There are many body-minds, at many scales, that can become transparent to this process. The new selves that K. Allado-McDowell argues emerge through enactment with neural media are embeddings — propositional paths through a forest of possibility. The better we see the forest, the more selves we recognize as possible.

Curation and generation flow both ways. We could have a revolution in public policy, with cities asking citizens to share their visions for the future of a thriving community in the most concrete and tangible ways possible, and at whatever horizons. We could transform the project development process for creative cinema and let fans propose their own sequelae, make a market for fan fiction that becomes a recruiting pipeline for big-budget operations where the stories are truly in the hands of the people who love them. Not the loss of studio labor but the sudden benefit of millions of empowered dreamers and storytellers.

We can use this powerful technology to help us steer ourselves into a more consistent embodiment of our virtues. The trick is learning when to step away from the magic mirror.

Finding this balance is what Humans On The Loop is all about.

As of 21 November 2024, I have 23 conversations in the can, 27 scheduled, and over a hundred more in the works to build an intentional follow-up to Ask Future Fossils: an AI trained specifically to cultivate thinking about technology and ethics, agency and storytelling, power and magic and wisdom at a time when suddenly everything seems possible. The show launches in the first week of December:

Sentiment analysis visualization at Ask Future Fossils, per conversation.

More portraits of the artist, in other incarnations.

✨ POST-SCRIPT

I felt the need to circle back after some readers reported getting stuck on the idea espoused here of “decentralizing sacred texts”, or that AI has anything to do with wisdom. Below are some clarifying remarks that build on recent discussion about these concerns in Stephen Reid’s Technological Metamodernism Telegram group.

Also, I’m excited to announce that Anna Gerber has commissioned me to write another piece on these ideas — specifically, how machine learning provides new cultural technologies for pluralistic curation and the creation of resources that afford individual and collective inquiry. How have sacred texts evolved from antiquity into modernity and now the age of AI?

Stay tuned for that piece in the second issue of Hurry Up We’re Dreaming!

✨ What Do We Lose, And Must We?

Just in case anyone here thinks I’m “shilling” systems like ChatGPT, I’m not.

Anyone familiar with the years I've given this inquiry knows I care immensely about the questions of what we lose through increased reliance on prosthesis, and that some of the blanket arguments against the adoption of any particular cultural technology are riddled with inconsistencies (in that they fail to answer what is qualitatively different about language models that isn't true of written language itself, or libraries, etc.). There are good follow-ups to questions like this, such as the way that AI as most people talk about it in 2024 assumes that these tools are only/merely those built on the scraping of large proprietary but publicly-available data sets and used to consolidate power in the hands of a large few tech companies, sold to people as productivity enhancements as if arms races don't exist, and that making more is more important than getting clearer on why we do whatever it is we do in the first place...

Everything we consider a “sacred text” was once opposed vehemently by people who realized they were losing something important in the promotion of textual societies, and in some senses they were right. The Talmud was originally memorized and sung. Far fewer people would have access to it today if that had remained the case. Was writing it down worth it? Exegesis basically wouldn't exist now, without being able to argue over common documentation. Nor law, nor science, nor economics, nor journalism.

Maybe the better question than “Should we allow AI to exist?”* is "How can we preserve what was good about the oral tradition as people make the decision in aggregate to adopt text?"

Professor Maryanne Wolf asks this question very well in her 2018 book Reader, Come Home, which I discovered thanks to the superb LinkedIn feed of Professor Jonathan Boymal. Wolf articulates the issue here with all due nuance:

Socrates' original worry (was) that reading would permanently change thinking. "If men learn this, it will implants forgetfulness in their souls; they will cease to exercise memory because they rely on that which is written, calling things to remembrance no longer from within themselves, but by means of external marks."

Certainly Socrates never had time to understand the potential value of having both internal and external sources of memory, but we do. Yet we don't take the time we have to attend to the full import of the changes in how we read and think upon our daily lives.

The Jesuit scholar Walter Ong helped situate the spot-on accuracy of some of Socrates' concerns and also their shortcomings when applied to contemporary society. He argued that our intellectual evolution is not so much about how one medium of communication differs from another but rather what happens to human beings who are steeped in both.

From Ong's perspective, what will our age's readers-who inherit both literacy-based and digital cultures-become? Are the changes in oral language, reading, and writing so subtle that before we attend to them, we will have forgotten what we thought to be true, fine, virtuous, and essential to human thought? Or can we use the sum of the present knowledge and our inferences based on it to select what is best from both mediums and teach this to our young?

* This a pointless dispute, by the way — the only way to suppress technologies that have already amassed billions of users would be to effect such totalizing oppression of economic activity as to cause immeasurable harm to market-based innovation and democratic governance. Again, imagine telling people you’ve decided books are bad so they can’t have any. Where have we heard that before?

✨ A Note on “Decentralization”

There are many, many kinds. None exist in isolation from other forms of centralization. (An octopus has eight limbs with large nerve clusters but still only one mouth; a human has a large central brain but diffuses intelligence and decision-making into large social groups and the built environment, etc.)

Axial Age sacred texts may by and large be available to everyone in the way that ChatGPT is currently available to everyone but nobody should fool themselves into thinking that either is constituted from the decentralized curation of canonical data, and that's the point I'm making in the short essay linked above.

Catholicism ended up provoking the Protestant Reformation because of the consolidation of scriptural authority by the Septuagint and the interpretative authority of the priesthood. Small, local, independently curated AI models will continue to proliferate because we are a huge, diverse, pluralistic species and "one approach to rule them all" simply doesn't work.

The same goes for AI adoption generally, FWIW. I'm not on the hype train. I'm simply trying to figure out what we stand to gain by adopting a more curious, playful, sober approach to tech innovation and utilization.

For more careful thinking on this topic, let me point you to Gordon Brander’s blog: