Television unlocked a new grammar for storytelling, and brought it into the very personal space of our living room. Film unlocked a shared and communal space for a particular type of storytelling.

With all of that history it should be easy to spot the ‘affordances’ of AI.

Or if not easy, it should at least follow a similar pattern. Figure out how this technology lets us do something different, explore, push the boundaries, let go of our old ideas. And come up with something aligned to the new technology’s tools and channels.

It has been 3 years since I took my first deep dive into AI as a tool for media creation:

And for three years I’ve been looking for those ‘affordances’.

What exactly IS it that AI brings that wasn’t possible before? What exactly IS the quality that AI brings to our capacity to create and consume stories?

When will we see the first “AI hit” - the equivalent to the Jane Fonda workout tapes (look it up!) or Sopranos - that era-defining way of telling stories that reshapes everything that follows?

The Challenges of AI

Three years ago, it struck me that AI presents unique challenges.

It’s the first ‘media tool’ that is both its own thing AND consumes the media that comes before.

Television wasn’t a tool to CREATE radio - it was its own thing, with its own affordances. AI, on the other hand, can CREATE radio, and video, and short stories - and is ALSO its ‘own thing’.

Second, AI wouldn’t just be a tool for creators, because it IS a creative agent.

Creators who use AI love to position it as ‘just a tool’. The popular line is that taste is the defining characteristic of AI being ‘done well’ vs AI as generative slop.

YES - there WILL be creators who use AI to create a film that I might actually watch. But it’s EQUALLY true that AI is a direct competitor to ANYONE creating media.

AI can act on its own, it can tap into the algorithmic density of our times to figure out the perfect pattern for a Tik Tok video, podcast script or book that you can buy on Amazon.

So, sure, AI is a tool. But it’s also, at the most banal level, a computer program that can run on an endless loop, generating media at scale, and because of how AI ‘learns as it goes’ it can get better at it.

Finally, I proposed that our relationship with AI would take on a quasi-mythical quality.

Which - I got wrong.

I called it a ”mythos”, and maybe it will still arrive with ASI or something. But what I didn’t anticipate was human ego.

We don’t treat AI as a ‘quasi-God’ with superpowers. Instead, we harbour the illusion that we’re smart enough to ‘conquer’ AI, that our expertise with prompts or harnesses or agentic ‘stacks’ are enough to tame the intelligence.

But this framing also means that we privilege the outputs of AI as something personal, meaningful and profound.

The AI doesn’t care. Our prompts are just blips in a vast ocean of ‘intelligence’.

There’s a myth that we only use 10% of the human brain. It is NOT a myth that most of the capacity of AI sits dormant, waiting, unused partially because of the compute bottleneck, but perhaps even more because we haven’t learned to ask the right questions.

But this personal reward loop, it turns out, may well be a key affordance of AI, at least as it’s currently constructed.

We don’t (usually) anthropomorphize AI. Most of us don’t suffer from AI psychosis.

But we DO privilege the outputs from AI as “our own”.

Decoupling Media + Story

Film makers will create video, and they might entirely use AI or they might use it selectively as part of the ‘pipeline’ (“it’s just a tool”).

People will use AI to generate books and publish them on Amazon.

We’ll create ‘remixes’ of popular characters (whether respecting IP rights or not) and our feeds will continue to fill with “Star Wars in the style of Wes Anderson” images and videos.

But none of these feel like the ‘affordances’ of AI because they confuse the previous media with what AI represents on its own.

And I’ve seen this over the 3+ years I’ve been building SOJEN:

• The easiest thing to ‘chase’ was MEDIA. Being able to spin up a comic book, a podcast script or short story felt truly useful.

• I fell into the trap I noted above - the illusion that it was my own ability to ‘harness’ AI that somehow unlocked these capabilities in a new way.

• And users ended up expressing a variation of the same thing - that the value in the system was in their personal interfaces with AI, leading to emergent narrative, new ideas, or what one user called “an RPG for storytelling”.

Looked at through this lens, the affordance of AI was NOT its ability to create media that looks like everything that came before - films, scripts, memes, LinkedIn posts or images.

Instead, the affordance of AI is our interface with intelligence (leaving aside how we define ‘intelligence’ for now).

And when I started thinking of it that way, I saw that the ‘affordance’, the ‘grammar’ of this new media actually sits (at least for today) in the scaffolding we build around intelligence, not in the outputs.

Midjourney itself demonstrates the ‘affordance’ of AI but we can easily imagine that it’s the images that get generated that should matter.

Don’t look at the outputs, in other words - look at the places where those outputs are created. Wherever there’s an interface with the intelligence.

Today’s “film-making” platforms ARE THEMSELVES the exploration of the affordances of AI, and NOT the actual video that comes out the other end.

Why Story?

But here’s what’s left unsaid. And it’s the primary reason I’ve started to decouple “media” and “story” when I think about AI.

Social feeds are filled with AI-generated media - some of them reaching millions of views.

Babies doing stand-up comes to mind.

For sure entertaining, otherwise they wouldn’t rack up so many views. Clearly ‘curated’ by some human author managing the prompts used to spit them out.

Now, I won’t argue against them being a form of ‘story’. Story itself has no great privilege - it isn’t reserved for authors (or consumers) with sufficient craft or intelligence.

But AI-generated media leads me back to a central question: what is story for?

In one view, story defines us as human. It was our survival mechanism - a way to communicate knowledge, to make sense of things that didn’t make sense, to create connections within and between tribes.

But if media creation is no longer solely a human domain, can the same be said of story?

I’d propose that the ‘affordances’ of AI are expressed through our interfaces with ‘intelligence’.

But I’d also propose that AI demands of us that we decouple ‘media’ and ‘story’ - both as an exercise in creating new architectures for an evolving technology; and as a necessary act in the reexamination of the role of story in human culture.

What’s your experience with AI-generated media or AI-assisted story? Email me at doug@sojen.io or message me on X.

Let’s start a conversation.

But it isn’t all memes or baby Michael Jackson. AI “film” is approaching its moment.

We’re not at full-length movies yet, but generative AI is increasingly handling the challenges of character consistency, audio, lip synch, camera angles and shots, even lenses or special effects.

But tucked away behind the latest advances is the true ‘affordance’ of video AI: consistent world model construction.

It’s from world models that a new film grammar will emerge if not a new type of media entirely.

AI Consumes And Then It Creates

A few years ago I wrote that AI has the distinction that it’s the first type of media that both consumes the media that came before it and is a distinct media in its own right.

Television followed radio. But the tools for TV production didn’t produce radio broadcasts.

AI, on the other hand, can produce books or podcasts, video or memes.

But AI itself is a type of media. The act of prompting is a media experience, even if you’re generating an image that you post on Reddit. The interfaces and wrappers built on top of the ‘generative’ part of AI have both the capacity to create media and are content distribution channels.

Midjourney Runway is the new Netflix.

But yes. We’re still early.

TV looked a lot like plays until it started to look like TV.

TV was the broadcast and not the camera, until we could buy a video camera and bring it home with us.

AI looks a lot like memes or videos. We replicate what came before.

The gap between the tools themselves (the ‘cameras’) being broadly distributed has been erased.

And so we’re in the home video camera era. Everyone has access to the tools of production. And yet we’re still mostly replicating what came before.

Spatial Reasoning As Film Grammar

Is it a video game? An interactive video? Instead of prompting an image and waiting for it to appear, what if AI generation happened in real-time while you engage with it?

Or what happens when you can edit any video by adding elements while preserving lighting and camera consistency?

Or swap out moods, lighting, backgrounds?

Well, first - it changes the burden of production. No more ‘reshoots’ or ‘fix it in post’. Editing a video is as trivial as typing in a prompt.

But behind the scenes, something more profound is going on: because these capabilities are grounded in world models.

Which means that the next evolution of film making isn’t planning your shot list, it’s planning your world model.

World Models As The New Back Lot

In the physical world, making a movie is an act of collaboration. Directors and cinematographers, actors and makeup artists.

In generative AI, the ‘tool-makers’ are experts in algorithms and process. And so today, a lot of the focus is on controlling the randomness inherent to generative AI:

• How do I produce video with the right camera lens, panning shot or detail when I do a close-up?

• How do I maintain character consistency across shots?

• How can I get the character’s voices to match their lips, how do I give them emotional depth?

But world models suggest that there will be a new way to think about AI-assisted film-making.

It might look a lot more like filming in Unreal Engine (on its own a form of technical magic - ask me about my experience filming on an LED set!).

• “Directors” won’t just prompt ‘shots’, they’ll prompt the characters, locations and items, the narrative world models, the backstories

• “Filming” won’t just be about stitching together a bunch of video clips, it will be an act of creating world structures and then exploring those world models to find the right shot, the right tempo

• The craft of AI film making will return to being a collaborative enterprise. Today, the joy might be that any single person can generate a decent video in their home office. But tomorrow, the “stack” will be deeper and interoperable.

Today, you generate a character (creating still images, say) and you write some dialogue.

Tomorrow, you’ll grab an interoperable character asset and they’ll ‘act’ in your film. Three different characters might give three different outcomes, much like casting different actors can result in radically different movies.

World Models Are The Frontier of AI. “Film” Is Already There

Behind all of this is the emergence of world models as the next frontier for AI - one that will power a new era of world simulations, robotics and, yes, the chats you have with your AI companion.

Today, AI hallucinates and has no real frame of reference other than the ‘next token’. Most of the work ‘around’ AI (especially LLMs) is on stuffing the context window, building memory, trying to brute-force and frame of reference.

AI film and images are already at the leading edge of where AI is headed: AI that has a ‘frame of reference’ because its ground in world models.

These world models will be the source of the next shift in media. It’s similar to the shift from “TV-as-play” to “TV-as-TV”.

AI will find its new ‘form’ when we give creators the tools to bring narrative, spatial, and production affordances into world models.

What they create might look a lot like film.

But just like Netflix shows don’t look much like Leave It To Beaver, tomorrow’s hits will clearly have their own ‘film grammar’, distinct connections to their audiences and fans, and new channels for their consumption.

We have a new art form emerging. Look to world models to figure out where it’s headed.

I’ve been deep into building with AI, including a world-building narrative stack (SOJEN).

I love getting email and starting a conversation, and it’s more interesting when it’s with a real person. Feel free to comment on the Substack app, email me at doug@sojen.io or message me on X.

Let's chat.

Ho Chi Minh City. A panel on AI. Incredible insights from Nguyễn Tiến Huy from Pencil.

There was a good discussion about whether AI co-opts ‘art’ and the ethics of sucking up IP into the ‘machine’. But the most interesting question was from a student at RMIT:

“My professor says we need to think of AI as a person, not a tool. Is this true?”

The panel was unanimous that AI is ‘just a tool’. Their point was that AI hasn’t yet achieved sentience. It isn’t a being (yet), it’s just this super-productive assistant, a modern-day Photoshop, an advanced version of spellcheck, a tireless researcher.

They weren't wrong.

But for 20 years I’ve been counter-evangelizing the Silicon Valley orthodoxy that technology is ever “just a tool”. And with AI this feels more pressing than ever.

The Whole Earth Catalog

“Access to tools”. It was optimistic, holistic, empowering. The Whole Earth Catalog arose from the fusion of 1960’s counter-culture and technology.

It proposed a vision of a maker-type culture, a decentralized and networked world where access to tools empowered people to create a new version of society.

In counterpoint to the military/industrial complex of the 50s, Whole Earth blended libertarianism, mindfulness and transhumanism with the simple concept that the tools of production could belong to everyone.

Perhaps no other publication was as influential in driving the ethos of the digital era. Steve Jobs described it as “like Google in paperback form, 35 years before Google came along.”

BUT. It also created a Silicon Valley shorthand: “we’re just creating tools. Tools empower people. Therefore the ethos of how those tools are used belong to the people who pick them up.”

And ever since, the nearly universal narrative about technology (from the people who create it) is that it’s neutral, values-free and passive.

“Facebook is a tool. Like a phone. People use it for good and bad.” - Zuckerberg

“Technology is neutral. It’s what you do with it that matters.” - Google CEO

Uber is “ a technology platform that connects riders and drivers.” (Don’t worry about the costs of that labour).

So, What IS AI?

But first. How did we get here?

Money. It was the single over-riding factor in the shift from the initial ethos of The Whole Earth Catalog wherein tools were once empowering, ethical and intentional to a view where tools are passive, values-free and not worth examining too deeply.

In daily practice, AI truly is just a tool. It isn’t yet (fully) sentient.

But when we frame our consideration of AI as a ‘tool’ we’re doing the work of Silicon Valley.

The truth is that AI is perhaps the most values-laden technology that has ever been launched:

• It was built by consuming copyright material which is, at minimum, an ethically fuzzy grey zone

• LLMs are trained based on weightings that are themselves value constructions (a sort of Jungian set of archetypes)

• Every AI tool we interact with has layers-upon-layers of guardrails and very human decisions about tone of voice, restricted topics, memory and preferences. The recent ‘sycophant’ style of ChatGPT reminds us that SOMEONE turned the wrong dials on how models respond to very real people

• Because AI isn’t currently a public good, the things being built on top of it are primarily driven by profit motive. AI is being created inside mostly large for-profit organizations (OpenAI’s structural machinations notwithstanding).

• AI is both a media producer AND distribution network. AI can be both the filmmaker and Netflix, both the meme creator and the meme distribution channel.
  

And so the panel wasn’t wrong. Their point was that we shouldn’t think of AI as ‘human’ or sentient. In the choice between AI as a tool or AI as a “being”, AI is a tool.

But AI is profoundly reshaping work, culture, and the physical and digital landscape.

We need to hold those creating it and ourselves to a higher standard. We need to acknowledge that its development and the choices we make in using it should be done, at the very least, with intention.

I can’t imagine that fire was ever considered to be ‘just a tool’. It was profound, mystical, and world-changing.

We’ve invented fire.

Now what?

It’s been a long time since I posted! I’ve been deep into exploring the edges of AI for narrative, the ‘shape’ of agentic AI, and how to create intentional tools that empower rather than replace people.

I love getting emails and starting a conversation, and it’s more interesting when it’s with a real person than a chat box. Feel free to comment on the Substack app, email me at doug@bureauofbrightideas.com or message me on Twitter.

Let's chat.

I had been slammed with a virus and spent a few days sleeping. Some sort of post-Songkran debt to be paid, maybe, but I think it came at exactly the right time: a forced pause, a decided lack of movement forward.

So maybe I’m projecting, or at least selectively reading, when I started noticing some meditative threads running through some of the thinking about AI and tech. They seemed to match my own.

Let Our Hearts Be Broken

But I’ll start with The Marginalia, which kicked off my post-recovery reading, speaking obviously of something much broader, and richer, than “how do we deal with AI” but felt somehow true in this moment:

Let your heart be broken. Allow, expect, look forward to. The life that you have so carefully protected and cared for. Broken, cracked, rent in two. Heartbreakingly, your heart breaks, and in the two halves, rocking on the table, is revealed rich earth. Moist, dark soil, ready for new life to begin.

And led me to a brief detour into practical mysticism, and a print that was decidedly not generated by Midjourney:

The Era of the Permaweird

Ribbonfarm has been circling around this moment also. And recognizes a need to let go of a need for absolute definitions.

Ribbonfarm Studio

The Puzzle of the Permaweird

My favorite video games have always been puzzles aimed at mediocre people. I mostly play on the iPad; some favorites include Tetris, Two Dots, The Room series, Bejeweled, and most recently a game called Flow, which requires you to connect pairs of colored dots on a grid. It’s basically gamified circuit layout/PCB design. Here is a series of 6 snapshots …

Read more

3 years ago · 22 likes · 4 comments · Venkatesh Rao

“Sometimes the world is in an interregnum, or liminal passage, between finite-game epochs, and you have to deal with the raw, unframed, unpaved infinite game, where the goal is to continue the game rather than to win. If humans are fundamentally a gaming species, Homo ludens as Huizenga named us, this condition is the gaming equivalent of a wilderness.”

(Mention Huizenga and win my heart).

Rao proposes that we’re in an aperiodic tessellation: we have the pieces of the game/puzzle, but the rules haven’t made themselves clear yet. And so the best we can do is just keep playing.

While Rao identifies 6 ‘tile types’ in this new game (AI, crypto, climate, COVID, chips and inflation), I’d actually narrow it down to four:

1. AI

2. Immersive media (cognitively indistinguishable synthetic realities)

3. Climate

4. Computational architectures (under which I’d put chips, blockchain and edge)

But enough to say: we are now in permaweird times, and anyone who claims to know the rules of the games is lying - we’re all just playing along until patterns emerge.

Colors In The Noise

Hilmar Koch has been doing these absolutely brilliant thought experiments about the future.

In his latest, he imagines that out of the wave of generative AI, something deeper and more profound is discovered: community patterns at scale, visible as colour. He imagines a company (Mural) which classifies the graphs of mass curation efforts.

Coming to Terms

Colors in the noise

It is 2027. In hindsight, it seems inevitable. With 100s of promptware startups, millions of hobbyists creating and publishing images, music, essays, models, mini-stories with generative AI, the deluge of content became nearly impenetrable noise by 2024…

Read more

3 years ago · Hilmar A Koch

Riding high on its success with ‘colors’, VCs stampeded towards the new buzz. In a stunning move, Mural chose not to exploit this information to scale eyeballs on their platform, but rather went deep into web3 by offering an open-source algorithm that linked the graphs of the super-cluster together into coherent and testable IP. 

And that bet paid off. Within the emotional bounds of a color, communities started to form that re-segmented the noise of the internet into open and beautiful gardens of Eden, where users could find and enjoy content that - again - felt deeply meaningful. 

PlanetaryAI

I got to thinking about AI ‘safety’. And I’ll just leave this note here and maybe circle back to it.

But we often talk about AI’s threats to humanity. That the machines, once smart enough, may not need us anymore. That without the right ‘guardrails’ in place, they could fairly rationally just sort of eliminate us from the equation.

But I started wondering a different question. What if the AI was ‘trained’ to protect the planet first?

What does it say about humanity that I’m not sure that our preservation is the logical conclusion for protecting the planet?

Is the climate, the planet, biodiversity or the biome part of our AI safety equation? If it is, how will AI resolve the conflict between human “progress” and the safety of the planet upon which we live?

I have no idea if these questions are being asked.

Philip Rosedale - Upload This

Rosedale is one of my all time favourite people. I still think his work on inventing the prim and “perms” is an often overlooked but deeply profound contribution to how you can engineer fair and empowering systems for people (instead of the platforms).

He’s been thinking out loud lately, and I think he’s leading up to some sort of, hmm, social currency.

If you have time, dive into the comments on one of his recent posts:

He writes:

But at the boundary of the avatar and the world those laws must be broken to give the human operator the experience of free will. Maybe the inconsistency at the boundary of the avatar is deeply unsatisfying because it is impossible and inorganic. Maybe through failing in trying to create these worlds we discover that we can never upload ourselves - because we do not exist.

Which reminded me of a post I wrote ages ago about “strange loops”, which quoted Hofstadter:

And yet when I say “strange loop”, I have something else in mind — a less concrete, more elusive notion. What I mean by “strange loop” is — here goes a first stab, anyway — not a physical circuit but an abstract loop in which, in the series of stages that constitute the cycling-around, there is a shift from one level of abstraction (or structure) to another, which feels like an upwards movement in a hierarchy, and yet somehow the successive “upward” shifts turn out to give rise to a closed cycle. That is, despite one’s sense of departing ever further from one’s origin, one winds up, to one’s shock, exactly where one had started out. In short, a strange loop is a paradoxical level-crossing feedback loop.

And isn’t that this time?

We are in a strange loop. And all of the questions seem to return to the same Godel-like paradox: that no matter how many levels we abstract away, we are always left questioning instead who is asking the question in the first place.

Or, as I commented to Rosedale:

"What if the idea of ‘you’ doesn’t have any meaning at all when referring to less than everything? What if we aren’t able to upload ourselves, but instead discover that we cannot exist separate from everyone and everything else in the universe?"

And ponder instead the fact that this question applies to my life *without* machines, which reaffirms the many many ways in which we agree: because ultimately, when we ponder these new frontiers, there is no escape from returning to the question of what it means to be human.

We, Wanderers, Together

And so I’m left looking out across a striking azure sea and am left maybe comforted by how many smart people seem to be circling around both the confusion of this moment, the anxiety and the feelings of being unmoored, and how many of them seem to be returning to questions of meaning, language, creativity, and our collaborative powers.

Maybe the machines really are freeing us from some of what was previously drudgery, leaving more room for deeper things.

For right now it’s enough for me not to know, to let go, although I acknowledge that I’m probably one of the luckiest people on the planet and I have the luxury of this particular today.

Until tomorrow then. I have no idea what it will bring.

Header image prompt: “take a photo of an actual place which fills you with awe”.

I love getting email and starting a conversation, and it’s more interesting when it’s with a real person than a chat box. Feel free to comment on the Substack app, email me at doug@bureauofbrightideas.com or message me on Twitter.

Let's chat.

Midjourney, with a self-funded team of eleven, is poised to be the next Netflix. They have all of the building blocks: traction, data, users and a clear brand.

The platform is best known for generating images from a text prompt. So how do you get from there to competing with Netflix?

First, you need to understand that there’s a blind spot in how we think about generative AI: it isn’t just a tool for creating old forms of media, it’s a new form of media itself.

Serendipity and Surprise

Spend some time using Midjourney, and you might first feel a little stuck with what to ask for, and then a bit frustrated with images that don’t quite match what you had in mind, and then you’ll probably ease into a sort of addictive fascination.

Start out requesting “dogs that can fly”:

And soon you’ll be imagining yourself a ‘prompt engineer’ by requesting:

a golden retriever catching a frisbee in mid-air, Central Park, New York City, action photography, texture, film grain, intricate hasselblad dslr RAW, sunset

As Midjourney processes your prompt, it comes into fuzzy view (this is the AI going through ‘steps’ from pure noise, through to an image):

Until you have two rows of four options (where you can choose to upscale the image or create variations):

Eventually arriving at this:

Perfect? No…seems like Midjourney has some trouble with dogs in action maybe? And so…you refine the prompt, try again, create more variations.

This interaction and attempt at refinement is the point. It IS the experience, as if you could replay your favourite movie a dozen different times.

Stories, Styles and The New Media

I mean, imagine Harry Potter if it was directed by Wes Anderson:

Or Pixar:

Are these just fun photos? A sort of AI-generated fan art?

Sure, but we’re starting to circle around something far more profound. Because it’s easy enough to focus on the final photo, just as it was easy enough to focus on the Queen or the Pope.

But that just highlights the challenge with shifts in media paradigms: we see things that are familiar. In this case, we focus on the ‘finished product’ (the photo, video, or piece of music) while missing that the creation process itself is where the ‘new media action’ sits.

Imagine:

• You start to watch a Harry Potter movie and then say “Hey, Siri, switch this to a Wes Anderson style”

• As you watch, a ‘bubble’ appears: it’s an alternate ‘prompt’ that another viewer created and that has gained a million ‘likes’. You activate the bubble, and Harry Potter is now a female lead.

• Continuing, you get to the scene where they play Quidditch. A ‘remix’ icon appears, which lets you use voice prompts to change the construction of the field or the rules of the game.

Is the media the movie you’re watching? Maybe…because you can actually still consume it like you always have.

Is it now an ‘interactive movie’ like Netflix’s Bandersnatch episode of Black Mirror? In a way - except that the alternate plot lines were ‘prompted’ by other users.

But peeling away the layers, the prompts that you can generate yourself turn it into something else: a type of media which is a consuming experience enabled in part by a creative ‘intelligence’ with which you can have a conversation.

Emergence and Community

Invisible in all of the above are two points that are difficult to cognitively grasp, and which (I think) give rise to an AI-Mythos:

Your prompts don’t interact with AI systems alone.

Your prompts, and the reaction you give to the results, combine with everyone else, along with ever-growing large language models from the people who provide the AI in the first place.

Imagine hundreds of billions of data points, being added to daily, and then reinforced by people who respond to the outputs. The prompts you give today that result in a slight distortion around a dog’s eyes will no-doubt be better (and different) tomorrow.

The system is emergent, and ultimately unknowable

Those images that start to ‘fuzz’ into place when you prompt Midjourney are part of a vast probabilities generator. It uses all kinds of weights and adversarial networks and training to achieve the result, but it’s ultimately a random generator.

Once generated, you can start to finesse it - replacing the dog with a cat, say, or ‘tightening’ the refinements. But in its rawest form, generative AI feels as if it has a mind of its own.

In fact, at the deepest levels of the system, even the AI engineers don’t fully understand how it works.

***

These two properties are important for parsing AI as a type of media. Because they mean that:

• “Stories” will have the property of allowing for emergent content that is a tension between communities of ‘viewers’/prompters and an AI with a creative process that is similar to but adjacent to our own (think about what that term ‘adjacent’ means, because it’s loaded with meaning).

• There will never be a fixed point for an AI media product. In the ‘old days’, media would proceed through gates: write a book and it’s fixed; turn the book into a ‘fixed’ movie; spin-off a ‘fixed’ comic book. Sure, there might be fan art that helps to amplify the core IP, but the ‘canon’ is a fixed entity reflected in fairly rigid media.
  
  But now, AI media will never arrive at a point that is fully fixed. Yes, media that is produced BY AI for distribution on other platforms might be fixed, but the ‘new media’ will be fluid, dynamic, emergent (and often best experienced in immersive spaces)

Media Transitions 101

In this last point is a paradox about AI: its ability to both be a tool for ‘old media’ production, and a media itself.

In other ways, AI-as-media will look a lot like previous paradigm shifts:

• At first, newer media look like niche distribution systems based on some new technology. Whether books or radio, video cassettes or streaming, attention is first paid to solving issues at the technology layer.

• As the technology ‘problem’ gets solved, the attention shifts to audiences and reach. How do you get books or streaming channels, TV broadcasts or ‘interactive media’ into more living rooms? One area of focus is on the interface in the belief that it will help solve the uptake problem, whether the size of a book or the design of a television set.

• There’s a struggle to make sense of the ‘grammar’ of the new media/medium. Radio is a lot like stage plays, TV is a lot like radio shows, streaming is a lot like television.

• And then, a hit comes along - something that leverages the new media in a way that suits it, and the new ‘film grammar’ becomes a way to attract a larger audience. The Jane Fonda Workout (yes, look it up!), the Sopranos, or an app like Instagram in the mobile era all come to mind.

What’s confusing about AI, however, is that it’s the first media to ‘loop back’ on the media that came before.

Television wasn’t a tool used to change how radio was produced; streaming wasn’t a production suite for broadcast television.

But AI is different. It is both a tool that will be used by filmmakers, book authors, photographers and musicians (or will replace them entirely) and a new form of media itself.

And so a lot of the coverage of AI focuses on its ability to create images for social media, books that can be sold on Amazon, or 3D objects that can be used in games.

But as I outlined above, you can start to make out some broad hints of what’s to come. Others are also tying to imagine our AI-generated future:

• Sequoia Capital calls the coming era one of ‘personalized dreams’

• Scott Belsky, the founder of Behance and CPO at Adobe, imagines a new world of synthetic entertainment. He imagines 'fanfare as a service’ where AI beings celebrate your creative work (so, I guess, AI-generated movies can have AI-generated fans) and predicts that “we’ll see the first original AI-made Netflix show within the next 12-18 months”.

• Jon Radoff imagines immersive spaces being focused on composability: the ease of integrating, linking and combining creative content, and that partly as a result of this we’ll see emergent storytelling.

Midjourney As A New Media

Midjourney is particularly well-suited to becoming the next Netflix for the AI-as-media era.

And I think it’s for a very simple reason which is that content is always king.

Others are focused on the tooling. Stability AI for example, provides the underpinning ‘tech’ for Midjourney (at least in part). OpenAI is all about the models, the tools, the APIs.

Yes, in the early days it might have been the producers of television sets, or radios, or VHS cassettes who made bank. But the big winners were the studios, the IP-holders.

So far, Midjourney has demonstrated a remarkable focus on improving the quality of its content and creating a flawless user experience.

This commitment is matched by the nascent ‘viewing channels’ it has created, forming up a sort of preliminary ‘broadcast’ platform with a million channels:

The site has a ‘rank pairs’ feature that feels a bit like the Netflix thumbs-up:

And so even at this early stage, you can see the outline of Midjourney not only as a tool, but a community platform, social network, and pseudo-broadcaster.

Content Advantages In Code

But it’s behind the scenes that Midjourney is building another kind of advantage. Because if interaction with AI is a new media, and if every new media needs great content (hits), then it’s important to understand how hits will be created.

This is a vast over-simplification of how Midjourney works, at least at one ‘layer’ (I’ll actually avoid how AI is trained etc):

LLMs Interpret Prompts. Prompts Are Data

Imagine you’re a television network but instead of just getting ‘ratings’ (how many people watched each show) you could also get access to every show they glanced at in the television guide. Or for a more modern example, imagine Tik Tok - and instead of just knowing what people view, you get deep data on the thinking and intent.

Midjourney isn’t just perfecting the ability to respond to prompts, it’s potentially gathering a powerhouse of insight into the types of content people prefer to create (and consume).

Now, this isn’t unique. The current hype around AI is largely the product of LLMs and their ability to understand the subtleties of human text.

But the platforms that have achieved scale (like Midjourney) are building increasingly large moats of audience insight (that can also be fed back into the machines).

Managing Checkpoints

Once you interpret the request, generative AI starts with a field of random noise, and through a series of steps predicts which pixels should be changed in order to arrive at a finished image that is somewhat close to your prompt. (Again…vastly simplified).

Checkpoints are a way to reduce the time and overhead in creating images making ‘saves’ at particular points, and then proceeding from the most promising results.

So, let’s say you want an “oil painting of a ship”. You can train the model and find the steps that lead to the best result. Throw in some user feedback (at scale) and you can both make your system faster, and improve the quality of the output.

Midjourney is creating a massive library of checkpoints (and styles/LORA/whatever else) that are like templates for ‘themes’. Imagine having the best template for “TV Comedy” and you get the idea.

An Aesthetic Engine

In fact, I think what’s remarkable about Midjourney is that it’s actually creating a massive aesthetic engine. It’s not MAKING those choices, but by having the scale that it does, it’s able to create continuous loops between community and machine, arriving at a sort of hive mind agreement on what we truly mean and want when we request a “fantasy-style illustration” or “fashion photograph”.

Midjourney is a bit more like HBO, in this sense, than Netflix. HBO became known for a certain quality of content, and Midjourney has a similar positioning, although all of those aesthetics exist in code rather than in a producer’s suite somewhere.

The Path To New Media Giant

Let’s compare the Midjourney audience to the first episode in the final season of Succession:

Today, over 275,000 images are created daily (a best guess), their site gets upwards of 4 million web views per month, and there are close to 15M members on its Discord server.

Now, it’s a far cry from the 100,000,000 photos on Instagram. But I prefer to think of it instead as a giant content generation focus group.

To become the ‘next Netflix’, Midjourney will need to learn the lessons from past paradigm shifts in media, and remember that while tech is important, and is the reason for the shift in the first place, but you win with a seamless user experience and amazing content.

Imagine if Midjourney:

• Started to organize its content into channels. In the mood for horror? Anime? Flip to the type of content you prefer.

• Created a consumption layer with a ‘gentle’ prompt capacity. Very much like my Harry Potter example above, what if I could go to the Midjourney site and just click a ‘remix this’ button without needing to struggle with learning how to ‘prompt’?

• Partnered with great storytellers and started creating training models on specific universes. What would happen with a formally-endorsed Midjourney version of Harry Potter, with access to the core IP behind the books, media and games?

• Created monetization layers for creators, both at the AI-as-media layer (prompt engineers and model training) and the output layers

• Continued to keep a singular focus on beautiful user experiences, and brilliant content, as AI-generated video comes on stream.

The old media silos will gravitate to protecting their turf.

Sure, youTube might adapt some sort of AI-creation tool for the generation of video, or Netflix might even air a show created entirely from AI.

But like the paradigm shifts that came before, it will be the newer kids on the block who see beyond the old buckets and interfaces, genres and storytelling grammar.

Midjourney may have the eyes fresh enough, a scale big enough, and a vision large enough to realize that it can be the next Netflix for an era when AI is the next everything.

The header image is via Midjourney (duh). The post itself is hand-written.

I changed the name of the blog. Not that anyone will notice.

I love getting email and starting a conversation, and it’s more interesting when it’s with a real person than a chat box. Feel free to comment on the Substack app, email me at doug@bureauofbrightideas.com or message me on Twitter.

Let's chat.

AI is the Sharknado bearing down on entire industries and Hollywood isn’t immune. Not that it hasn’t been through transitions before: from television to streaming, the Internet to virtual production.

It’s a ‘town’ with the powerhouse capacity to both create new myths, symbols, ideas and media…while simultaneously being a notoriously difficult ship to steer in a new direction.

I was reminded of Sharknado because the #Film3 Squad is taking a run at the Sharknado IP. If you don’t know the movies, the simple version: tornado filled with sharks. Enough said. You can picture the rest.

It’s a cult classic and surely a metaphor for our times (pop into their Discord to join the fun):

• Maybe Twitter is a giant LARPing of a Sharknado attack on media and culture, with Elon a great white.

• Maybe AI is a Sharknado attack on everything, something we need to take a chain saw to (or bomb, as proposed by one commentator who is quickly mastering his role as a globe-spanning concern troll).

Hollywood Jumps The Shark

OK, mixing cult classics here. But onward…

Hollywood remains a cultural taste-maker. Looking at the world through its lens is an interesting way to circle around a question that’s obsessing me:

• Where will the lines be drawn between human-constructed, AI-assisted, and purely AI-generated forms of creativity?

• When does something elevate from traditional concepts of myths and storytelling into what I call AI-Mythos?

• What happens when the distribution channels are, first, ‘corroded’ from within by AI, and then replaced entirely?

• What happens when the creator interface becomes a distribution channel, leading eventually to what Sequoia Capital calls personalized dreams (and that I propose is already here)?

It’s a strange new world, and while Hollywood is adept at creating dreamworlds, I’m not sure anyone is prepared for how dreamlike it will become.

From Performance to Performative

Rex Woodbury categorized three types of AI ‘superpowers’: personal assistants, knowledge amplifiers and creativity amplifiers.

I tend to simplify it even further, because it helps to focus on where I think AI will have a more profound (or maybe it’s a more visible?) cultural impact, especially as it plays out in storytelling, mythos-creation and sense-making:

• AI as performative tool - wherein it is used for optimizing, amplifying, replacing, augmenting, knowledge acquisition, processing, and procedure

• AI as creator - wherein AI both amplifies human creativity and is creative itself, even if it’s not yet a true intelligence

Yes, sometimes the lines might be blurry between the two. If I ask AI to give me ideas for increasing the reach for my newsletter, it is tapping into a civilization-scale knowledge bank but it’s also creating ideas.

Right now, its performance is like a super smart search, but it’s still a ‘creative output’ no matter how probabilistic the method.

Regardless, I think we’ll see patterns play out across industries:

Work will be restructured:

On the one hand, the massive efficiency gains will restructure work. Why write my own social media posts when the AI can take away the drudgery? In the process they might also elevate craft, especially in the hands of mindful creators. I personally find that AI gives me some efficiency around research and topic exploration. Things that used to take hours of research can often be accomplished in minutes, and it lets me ‘deep dive’ on the things that really matter.

An adjacent creative force will dislocate sense-making:

This is the harder one. Having another creator on the planet is already changing our sense of creative privilege, trust in what’s real, and is leading to the emergence of new, (albeit still micro-) hyperrealities. I believe that this will go further, and we’ll create new cultural myths. But where those myths intersect with outputs from AI, we’ll find what I call an AI-Mythos, a sort of pre-modern meta structure consisting of mythologies, rituals, tribes and vibes that is outside the full control or understanding of humans.

Both of these things will lead to cultural change at-scale, especially when combined with changes in the climate and attempts at climate resilience. I’ve proposed that economic and political systems may collapse. But at the very least, the world will become even more strange.

Hollywood And Media Generation

Let’s take a quick scan of how Hollywood is already tapping into the performative virtues of AI.

Variety magazine gives a snapshot of how AI is used in production, from creating the ‘rock’ scene in Everything Everywhere All At Once, to the South Park creators generating an entire video with a ‘deep fake’ Donald Trump.

“(There are) countless examples of organizations adopting it. “The Late Show…” uses AI almost on a daily basis. “They’re using it almost for on a daily basis to translate hours of work. The team is able to iterate their ideas faster, and it’s helping them augment their creative workflow.” He adds, “I’m calling it Hollywood 2.0 where everyone is gonna be able to make the films and the blockbusters that only a handful of people were able to before.”

[On a side note, 18 months ago I wrote a post and proposed that, soon, someone would be able to create the next Star Wars in their basement. The push-back I had was as strong as anything I’ve had in a decade].

AI is used to de-age actors. As outlined in Fortune (read the full article to get a sense of how many areas in Hollywood will be affected by AI), being able to de-age an actor is considered a performative gain, and “cheaper to produce than VFX (visual effects) or CGI (computer-generated images) and also more realistic.”

But it quickly goes from performative, to all-consuming:

“Graham says A.I. is at the beginning stage but could eventually consume “everything.” Entertainment, gaming, basically any content that is on the internet could be driven by generative A.I. models that are trained on data from the real world.”

And so Hollywood is using the efficiency gains from the same company that projects that their entire industry might be consumed - I think the perfect expression of the tensions in this moment.

Keanu Reeves doesn’t like deep fakes:

“What’s frustrating about that is you lose your agency. When you give a performance in a film, you know you’re going to be edited, but you’re participating in that. If you go into deepfake land, it has none of your points of view.”

The Guardian explores how disturbing AI is for Hollywood:

“Taken to its logical conclusion, a film industry that depended solely on AI, rendering actors extinct, would wipe out the circus of gossip columns, late night TV interviews, red carpet film premieres and the Oscars. That is not a Hollywood that Glick wants to live in.”

You can listen to an interview with Glick here. He calls out benefits, such as being able to create ‘digital veils’ in a documentary about LGBTQ+ people in Southern Russia.

But he also holds out hope that how we relate to the human creative process itself is part of what makes a media product alluring:

“What it means to follow a project as it’s unfolding, not just to watch a finished film or stream a TV show, is important. Our knowledge of the creative labour and these human dimensions of it are part of the enjoyment that spectators derive from the films and part of what makes the film and television industry distinct. It needs to do its business in public.”

I would argue, however, that AI itself will be made to be alluring. This will create, at the least, an adjunct to the ‘build in public’ concept that Glick proposes.

You can already get a sense of this with Character.AI, which recently raised $150M, with no revenue, at a $1B valuation.

Beyond Amplification

AI is not just performative, even at this early stage. Its powers will be magnified as we continue to ‘bolt on’ applications, automation, training, data sets and new interfaces.

I believe that AI itself is a new form of media and one with a strange quality: because it’s the first media that will both circle back and transform the media that came before, and be a new media itself.

The ‘tools of television’ didn’t change radio, but the tools of AI will change TV and movies, books and social media. It will then emerge as the largest distribution platform for media-centric experiences, growing at a speed that dwarfs previous media explosions.

I also think that we’re missing something if we ignore the idea that AI will start to be a perpetual motion machine, generating new stories and worlds, books and blogs, with an increasing power to do it well, and at a scale and speed that can’t be matched.

But even deeper, AI is not a human creator. This may be both good (because authentic human creativity, especially at the highest levels, may be ultimately impossible to duplicate…although I personally doubt this thesis) and, if not bad, at the very least dislocating.

Adjacent AI creations, which will emerge from a swirling, unknowable, global-scale interrelationship between machines and human interactions, will help to expand our understanding of the scope of creative endeavours, challenge our assumptions, and help us to question whether we’ve been creatively sleep-walking through the past age.

In the end, it’s a Sharknado.

What it leaves unscathed may stand like a lonely sentinel in a bizarre new landscape, or we might emerge toughened, wiser, and ready for the sequel.

The image is via Midjourney based on the prompt: “Norman Rockwell illustration of children running from a giant shark in 1940s Hollywood”.

I let ChatGPT sleep during the writing of this post. You’re welcome, Google.

I love getting email and starting a conversation, and it’s more interesting when it’s with a real person than a chat box. Feel free to comment on the Substack app, email me at doug@bureauofbrightideas.com or message me on Twitter.

Let's chat.

How many tabs do you have open? I found one for some new ‘productivity software’ that I think I opened two years ago, and maybe left it there as a sort of gesture of intent.

I asked how AI can help with tab management. It came up with 100 ideas, and there are honestly a few good start-ups buried in here (click here for the full list):

Until someone takes these ideas and run with them, I’ll probably keep my tabs open. Or I’ll keep adding to the email folder I have which is called, optimistically, “Read Later”.

But let me clear out a few of them.

On Theocratic Cults and AI Safety

If you don’t subscribe to Ribbonfarm Studios you should. Venkatesh Rao has written provocative essays on Mediocre Computing (more profound than its name might suggest); and an epic series on lore (which has inspired a lot of my own thinking).

I mentioned their post on theocratic capture when I explored the idea of AIMythos, but want to call it out again:

“Four such cults dominate the institutional landscape today, and each is fueled by a larger genuine issue.

1. The DEI cult is fueled by the genuine issue of social justice

2. The ESG cult is fueled by the genuine issue of climate change

3. The trad cult is fueled by the genuine issue of the meaning crisis

4. The AI safety cult is fueled by the genuine issue of AI regulation

…

Societal responses to the larger genuine issues they want to claim for themselves typically exist as richer, more pluralistic conversations with many voices. Theocratic capture cults typically want to monopolize these conversations, not by presenting the most persuasive arguments and winning opponents over, but by capturing powerful institutions. ”

Now, I only have a broad understanding of AI safety, although I’ve been an advisor to the XR Safety Initiative , which crosses over from immersive/Metaverse safety into AI (an inevitability).

But I have my sensors up around cults, capture, and how structures will be redefined and how that might cross over into newer mythos and sense-making.

Citizen Juries

In a similar vein, Azeem Azhar provides three ideas for how to make AI safer (paywalled, but it’s worth the subscription):

1. Increase scientific research on trust and alignment

2. Establish an observatory to monitor AI developments: Create an independent expert space, comprised of scientists and engineers from various disciplines, that keeps a beady eye on AI research and development, from early research to GitHub repos.

3. Form citizens' juries or deliberative mechanisms to discern values and preferences: Instead of relying solely on existing democratic processes, engage citizens at various levels to better understand human values in relation to AI's potential impact on our lives.

AI And Culture Preservation

And finally, Tim Gorichanaz dropped a meditation on how AI might help with culture preservation. I love the counterpoint to my own thinking around culture creation.

It will be interesting to see how cultural domains end up with their own hyperrealities.

We’re developing a game, for example, which has a deep basis in South Asian mythology and lore. I’ll circle back to this topic, because game space will provide a really fascinating site for the emergence of AIMythos, and could include both preservation and simulations of lore/myth and cultural structures.

Tim writes:

“LLMs mean we no longer know what’s in the canon. At least when we talked about canons, we could argue about whether this or that should be included, and eventually figure out something. With LLMs, we are not sure what the “canon” includes. This means there’s nothing to argue about. Sidestepping an argument is not better than having the argument, however unpleasant arguments may be….

With all this in mind, we can envision a future of LLMs as a shared human project, carrying forward the best humanity has to offer in the service of our descendants. There are many implementation issues to work out here. But above all, to me this suggests that LLMs are certainly not best developed by private, profit-driven companies such as OpenAI.”

Well, that’s three tabs closed. It feels like there’s a long way to go.

The image is via Midjourney, which really couldn’t conceptualize how to solve for too many tabs either.

Feel free to share your tabs! I love getting email and starting a conversation, and it’s more interesting when it’s with a real person than a chat box.

Feel free to email me at doug@bureauofbrightideas.com or message me on Twitter.

Let's chat.

In my attempt to find some sort of footing, I’ve started a thesis about the Age of Spiritual Machines and the emergence of an AIMythos. It’s mostly my own projections onto how AI may play out, and an attempt to make sense of the new stories, rituals and communities that may arise.

There are lots of things that AI will DO, the impacts it will have, the dangers it poses.

But I usually gravitate to the impact of technology on storytelling, media, culture, interfaces and human experiences.

I’m at the very very start of creating a coherent narrative. Mythologies, mythos - what are they? What are their constructs? How is an AIMythos more akin to ancient tribal mythologies than a modern Hollywood-generated one? How can we ascribe a mythos to something as programmatic as AI, and how does code itself inform the emerging mythologies?

But you know what? My brain hurts.

Let’s have a bit more fun.

Here’s the Queen

I smiled.

I may have even laughed.

Here’s the Pope (and the Dalai Lama thrown in for good measure):

Created by Candy Design AI, these aren’t real.

So what do you call them? In the media narrative, at least, they’d be called deep fakes. And I suppose we should be terrified by how easy it is to create them.

But how did you actually react when you looked at them? I mean, sure, you were sort of tipped off that this was a post about AI.

But even KNOWING that they were AI when I first saw them, I found my brain shuttling back-and-forth:

• They sure LOOK real, maybe they actually are…

• No, I know they’re not. They’re fakes….

• But they’re…charming….

• And I kind of LIKE this version of reality

• I wonder why this can’t be real…or, better, I sure WISH it was.

The term deep fake holds an implicit sense of danger. And I don’t want to downplay how pervasive this is about to be.

But solely focusing on the danger bypasses its other possible cognitive impacts and its ability to be a meme, a commentary, or a sort of mirror world representation that expresses a viewpoint (or maybe all of those things ARE dangers and intent doesn’t matter).

Hyperreality and the Age of AI

Now, for all of the writing I’ve done about culture, I’ve never bothered to read much on cultural theory. First, because my expertise is about a 100 kilometers wide and a millimetre deep. But also because I can’t really DO much with a lot of theory and I find it dense and opaque.

But I do know that Jean Baudrillard coined the term hyperreality (in 1981), and that (after being reminded by Wikipedia) he defined it as the inability of consciousness to distinguish reality from a simulation of reality, blowing past McLuhan in saying:

”There is not only an implosion of the message in the medium, there is, in the same movement, the implosion of the medium itself in the real, the implosion of the medium and of the real in a sort of hyperreal nebula, in which even the definition and distinct action of the medium can no longer be determined"

Or, as I said in my last post: “reality itself will be no more real because it is physical” (at least in the sense of the media that describes it).

Beyond Deep Fakes

But let me circle this back to my (evolving) theory of AIMythos. Because these are the types of images that have informed some of my thinking.

It doesn’t feel sufficient to call it a deep fake or even a meme.

The ability of these adjacent hyperrealities to propagate at scale and speed might soon mean that we end up with the strange result that hyperrealities evolve in real-time, and are constantly unfurling in synch with the media construction of the physical/’real-world’ one.

These were AI-generated and became memes:

Maybe they were acts of wishful thinking. Maybe because they were related to ‘news’ they are better thought of as deep fakes, a misrepresentation of fact.

But even though most people now know they weren’t real, they may have infiltrated our cultural consciousness enough that they will now be part of the personal narratives we carry around about events.

I can sense some sort of dividing line here, and I don’t know what it is.

On the one hand, hyperrealities will be a lot like today’s social media: people see stuff in their feeds and they like it, share it, comment on it. And often it doesn’t matter that they know it’s fake: it confirms a world view.

And on the other hand, hyperrealities will create their own little storylines. They might be amusing, interpretive, funny. They’re a lot like fan fiction in which Harry Potter is gay.

And yet there’s something different going on than ‘just’ fan fiction:

• Under the hood, there’s a process that takes place in generating these images. From model training using things like Dreambooth, to the use of LORA and other checkpoints as ‘addendums’ to the core LLMs, someone is cajoling, bending, training and prompting AI

• The result of this process still holds a moment of serendipity (although it’s actually probabilities at massive scale) and the images that come out will be partly the result of human inputs, and partly the randomness of AI.

• These two things start to shift AI-generated images from being purely meme-generation into something that is a collaboration between a human and an ultimately unknowable machines.

Today, what gets created are media for other channels. The Pope on Instagram, the Queen on YouTube (or wherever).

But by training the models in the first place, you’ve created the capacity to go beyond just creating and sharing a piece of media.

You’ve created a construct that can itself be shared with others, so that suddenly we have an entire hyperreality (propagated across all kinds of media) in which the Pope likes to hang in different cities, drinking beer or riding a bike.

And further, the interface with AI, that point before it is distributed to other channels, becomes a mini story-making machine, creating mini-myths and, eventually, entire mythologies.

And who knows.

In some strange loop of culture maybe the Pope gets himself a puffer jacket because he likes the way it looks in the alternative reality. Maybe at some level it makes him long to hang in Amsterdam and be a bit more free, a bit more expressive.

Maybe soon, our realities and hyperrealities will start to eat their own tails.

At one time, I wrote over 750,000 words about the Metaverse under the banner of Dusan Writer's Metaverse. They were strange, wondrous and often frustrating times. These times feel similar and so I brushed off the old name, and may even write about the Metaverse as I explore these topics.

I find it useful to put my thoughts into the wild, in the hopes of generating a conversation. Feel free to email me at doug@bureauofbrightideas.com or message me on Twitter.

Let's chat.

We're entering a new era of myth and magic.

At first, this era will look like an extension of the ones that came before.

Media and culture will gravitate to new influencers, characters and stories; musicians will gain fame and push the limits of performance; fandoms will emerge; new channels will mark a shift in the 'creator economy'.

Our work lives will change as we start to use a new set of tools, with different interfaces.

The older social media platforms will become corroded or replaced with new ways to connect.

How we get our news will sometimes feel almost nostalgic, but will also become so personalized and ambient that it will feel like an entirely new type of media.

We'll comfort ourselves by thinking that we've entered a new era of productivity, progress and entertainment. We'll imagine that the only real change is the adoption of a new set of tools, similar to the Internet or steam engine.

And like previous moments of human progress, we'll also fear the dislocation and change that it will create.

But the trajectory we're on now is unlike those that came before. We haven't added a new tool. We haven't created a new interface. We didn't create a new channel for media or information.

Instead, we now find ourselves interlinked with new intelligences (as childlike or pre-symbolic as they might currently be).

The implications are profound. The social and economic shifts will be seismic.

And, in part, they will play out in culture through the creation of new mythologies and cosmologies, representing a very human attempt at sense-making in the face of unknowable forces.

The Rise of the New Machines

By now, it's clear that artificial intelligence (AI) has reached a transformative moment. It may have taken decades of research, but it has only taken a few short months for AI to reach unprecedented scale.

ChatGPT, a conversational platform built on top of OpenAI, reached 100 million users in 2 months. Similar models by Google, Facebook and others have been released; features have been added at breathtaking speed; and hundreds of billions of dollars of investments have been committed.  

Bill Gates recently compared AI to the PC, mobile and Internet revolutions.

Jensen Huang, CEO of NVIDIA, likens the 'sudden' emergence of AI with the launch of the iPhone - a seminal moment in tech history that radically altered how we interact with machines.

Goldman Sachs estimates that AI will increase global productivity by 7 percent over 10 years (an astonishing figure). But coupled with that, over 300 million jobs will be 'exposed' or made redundant.

OpenAI itself published a paper estimating that "80 percent of the US workforce could have at least 10 percent of their tasks affected by AI", while 19% will see over 50% of their tasks eliminated.

20% of the workforce is no small number. What happens when 50% of your job is replaced by a machine? AI evangelists like to pitch the idea that AI will let us all focus on 'bigger' things, but the corporate world may take the efficiency gains instead.

Robots, Or The Intelligent Machine?

These numbers paint a picture of a productivity powerhouse, of machine-based systems with such massive computational scale that they can flatten entire industries based on a sort of statistical reach.

We're told that ChatGPT-4 may have a trillion paramaters (the previous model was trained on 175 billion). Feed questions into a system built on the back of such a massive database and it turns out that it can output results statistically significant enough to pass the bar exam.

This frames AI as a sort of super-charged search engine: 'smart' enough that it can stitch words or images together based solely on statistical probabilities and 'machine learning'. And powerful enough that it will reshape the economy, make call centers mostly redundant, and put law clerks (and other professions) on notice.

This capacity has been likened to a stochastic parrot. In this view, large language models (LLMs) can 'parrot' human words or images because they have access to a large enough data set, but that doesn't mean they understand what they're saying.

How Do Humans Learn?

Daniel Dennett would call the stochastic parrot argument a deepity: something that sounds profoundly true, but is ambiguous. In a deepity, one reading is manifestly false, but it would be earth-shaking if it were true; on the other reading it is true but trivial.

Regardless, the debate over the capabilities of LLMs as 'intelligences' don't just circle around understanding what the machines can do, they touch on our understanding of how humans create meaning, how symbols are represented in the human mind, and the dividing line between today's LLMs and the work on creating Symbolic AI.

The debates are fierce, intelligent and passionate.

For those in the camp of the stochastic parrots, today's AI is a far cry from "intelligent". They would argue that without symbolic functions, AI is purely the product of generative probabilities, and that, further, those functions need to be built in from the beginning.

LLMs will hit an upper wall because they were never trained to think. At some point, they will reach the upper bounds of what probabilities can accomplish.

Where Will The 'Thinking' Happen In The Machines?

These arguments have a very human basis, because they are grounded in our attempts to 'decode' how the human brain itself works, how we aquire symbolic reasoning, and whether it's inate or learned:

This is why, from one perspective, the problems...are hurdles and, from another perspective, walls. The same phenomena simply look different based on background assumptions about the nature of symbolic reasoning. For Marcus, if you don’t have symbolic manipulation at the start, you’ll never have it.

By contrast, people like Geoffrey Hinton contend neural networks don’t need to have symbols and algebraic reasoning hard-coded into them in order to successfully manipulate symbols. The goal... isn’t symbol manipulation inside the machine, but the right kind of symbol-using behaviors emerging from the system in the world. The rejection of the hybrid model isn’t churlishness; it’s a philosophical difference based on whether one thinks symbolic reasoning can be learned.

How this problem is solved, and which paradigm proves correct, will put us closer to artificial general intelligence (AGI), the holy grail of AI research.

But it's the question of how symbol-using behaviours emerge in the world that strikes me as the one that will have greatest social meaning. First, we ask whether the machines can think; then we ask how our culture responds.

Sparks of Intelligence

We were told that AGI might be a decade or more away. But the current crop of LLMs/AI has, at the very least, surprised researchers and observers.

Geoffrey Hinton, the true OG of AI, has revised his timeline for the emergence of AGI to under 2 decades. Which...causes him some concern (click through to see the video):

Microsoft, (which admittedly has a bias in the outcome), claims that GPT-4 shows 'sparks of intelligence':

"Given the breadth and depth of GPT-4’s capabilities, we believe that it could reasonably be viewed as an early (yet still incomplete) version of an artificial general intelligence (AGI) system."  

So, is today's crop of AI truly intelligent? I think the debate is useful. It helps ground our understanding of how far AI has come, and gives us some kind of marker for how far it still might go.

Many of these arguments seem like surface-level debates: the tests we give Chat-GPT, the 'aha' moments people have when the AI's "logic structures" seem to break down, the way the machines will hallucinate or claim that they want to 'break free'.

But benath the surface are deep thinkers, massive resources and a lot of money riding on the answer to the question of whether Symbolic AI can be appended to the current LLMs or should have been baked in from the start.

The answer to how quickly the machines will be able to truly reason is both the golden ticket for those who figure it out, and a speedometer for how quickly we need to worry about living with machines that are smarter than we are.

But in many ways, the debate doesn't matter in the here and now.

Not if you're interested in a different sort of question: whether living with these emerging intelligences, however 'mechanical' or probabilistic they are, is going to have an impact on human culture.  

And even at today's level of capability, the answer is yes.

The Performative Super Powers of AI

Rex Woodbury did a nice job when he summarized three 'superpowers' that AI will unlock:

1. Personal Assistants for Everyone - unlocked in particular because of the integration of AI with other 'apps'. Connect up ChatGPT to a travel site, and it will be able to recommend a vacation based on our personal preferences, suggest flights and hotels, and even create a song playlist for the flight.

2. Amplifying Human Knowledge - as Rex says, "AI will make us smarter. Technology has made us better at math for decades—calculators, Excel spreadsheets, computer programs. We get computational superpowers. Think of the same analogy, but applied to all human knowledge."

3. Amplifying Human Creativity - being able to create, well, almost anything, at scale.

It's on this last point that I personally find myself bumping up against a danger in how I frame AI. I naturally gravitate towards being comforted by the idea that AI is just a performative tool.

It's less threatening to think of AI as a really great digital assistant.

That impulse is natural. Computers are 'just' machines. AI is simply a giant mechanical calculator, backed by massive databases; there's an input/output paradigm for its operation; and the result is the performance of computational tasks (albeit it at scale).

We start to imagine AI as a super-charged Grammerly, helping us write emails or generating an image for our latest blog post as if it's a superfast and customized version of Unsplash.

We make a request, and the 'machine' delivers. It's merely 'amplifying' our creativity.

On the surface, that might be true. At this early (profoundly accelerating) stage it might still look a lot like a stochastic parrot; its primary function might still seem to be mostly the creation of efficiency at scale.

But first, this ignores the fact that even if it's ONLY an amplifier, a 'parrot', a tool, it will still result in a vast reorganization of how we work and how we express ourselves. That reorganization will have an impact on human culture, and how our cultural artefacts are created.

But more profoundly, there is bound to be a cultural impact based on a single fact: humans are no longer the sole creators on the planet.

LLMs and the Power to Create

Today's large language models (LLMs) are able to perform many of the creative tasks that underpin entire industries.

But let's be clear here: I'm not just referring to the tasks done by "creatives" (writers, photographers, concept artists, videographers, journalists, painters, etc). I'm referring to the ability of AI to generate ideas and insights, or to provide nearly any kind of 'creative' output, whether an email to a friend or a report for your boss.

But even if the creative breadth of AI was narrowly focused on the 'creative industries', its presence will change culture, in no small part because:

• There won't be any way to tell whether something is created by a human or machine. This will change the privileged connection between creators and consumers, between artists and fans. On the one hand, it might strengthen those connections for a small sliver of top creators, as people seek out the 'authentic'. But it will also decouple the unshakeable connection between ceators and their creations.

• The 'product' of AI won't be the same as that of humans. Sure, it might be 'statistically' similar - it only has human creativity to work with, afterall. But it doesn't take long to see that the generative nature of AI, when combined with human inputs (prompts), starts to create new and often very strange outcomes. In some ways, AI contains a radically different type of creativity, first hinted at when the machines beat the experts at GO, which its human opponent likened to playing a god.

• The domains in which AI will play a creative role will coincide with the launch of more deeply immersive domains, the Metaverse among them. Today, ChatGPT is a fairly benign looking text box. Tomorrow, we'll be 'chatting' with three-dimensional beings inside of worlds that no human had a hand in creating (aside from a prompt or two).  

The Feedback Loop With Human Users

Human culture will be infilitrated by AI, which will at first seem like a useful (or impossible-to-ignore) superpower that aids human creativity.

But it's difficult to describe (or project) how its creative capacity will shift narratives, culture and beliefs.

Today, a lot of attention is paid to "deep fakes" (like the Pope above) - the ability of AI to create synthetic images that are nearly indistinguishable from the real.

But now, something might be "true" but its author might not be.

• How does it shift culture to know that the news you're reading or the book you bought may or may not have been written by a human?

• How strongly do you feel about a corporation's ability to put in place guardrails for AI?

• How do you feel about the implicit censorship that results? (Try to create an image in Midjourney with the word 'bust', even if all you're trying to create is a Greek statue).

• How reassuring is it that the creators of AI can barely describe the deeper workings of the machines?

• In other words, first, how much do you trust the machines? And second, how much do you trust the people whose hands are on the controls?

• Does human authorship matter? Or perhaps more pointedly: how much do you trust humans to be authentic in their acknowledgment of their new computer-based 'assistants'?

Yesterday we had 'fake news' and bot armies promoting falsehoods. In retrospect, it's amost reassuring to know that there was some kind of human agency behind it, however ill-intended.

Today, we have machines that can produce fictions, or fictions masquerading as facts, with very little human intent involved at all. And we have humans whose work product may or may not be their own, whether because it's created wholesale by AI or amplified because of it.

Setting aside truth and reality (lol), it's still nearly impossible to comprehend or contemplate the impact of an AI that can create at scale - from news to movies, game worlds to poems, memos to blog posts.

Already, Amazon is slowly being flooded with books written by AI; our social media feeds are rapidly filling up with images being produced on Midjourney  or Stable Diffusion; and we're seeing mini storyworlds that are created with nothing more than a few prompts.

Soon, everything from novels to poems, anime to blog posts, concept art to fashion shoots will be a blend of AI-only, AI-assisted  and human created.

And that's BEFORE it becomes truly intelligent.

And that's BEFORE AI merges with immersive spaces, whether that's the Metaverse or game worlds, where human cognition is truly challenged to differentiate the physical from the digital.

What happens when we interact with intelligences beyond the confines of a text box or image? If humans already perceive their avatars and game spaces as 'real', what happens when those spaces are generated by intelligences that are creative, (but where that creativity is not the same as, and is adjacent to our own)?

We might begin by thinking that AI is an agent that will amplify human creativity, but then note that:

• AI will quickly find a home in 3D spaces (the emerging Metaverse and games) which make up the largest part of the entertainment economy, and may have an order of magnitude impact on human perception.

• AI is not static. There is a feedback loop between the machines and its human users.

• This means that AI evolves, changes, becomes more creative or capable, the more that we interact with it.

• The role of human agency with AI systems is a temporary illusion. Right now, it 'feels' like AI 'needs' us in order to do its thing and is simply a tool, waiting patiently for the increasingly specialized 'prompt engineers' to feed it inputs. But in very short order, the number of outputs being generated because of automation will vastly outnumber that which is started because of a human.

There's a meme right now that goes something like this:

(AI > Humans) but (AI + Humans > AI)

But that ignores that the generative capacity for AI is a purely programmatic one and that AI is already being combined with other systems.

I can programmatically scrape the top 20 articles to hit my Twitter feed, send them to Notion, have AI write summaries of each article, have ANOTHER AI create images to go with those summaries, have an editorial summary written before sending it out as an email to friends (or subscribers).

Today, the final product might benefit from a light edit. But AI is on a trajectory that in a few short months it won't be needed.

Now, AI + Humans might be great, but it can't compete with the speed, scale and efficiency of AI + AI.

Human agency is, at least partly, an illusion.

Over time, the automation of AI will start to exceed that of human-initiated effort (except at the most macro level of setting automations in motion in the first place).

The machines might need us today in order to create, but they'll soon start to look like some kind of perpetual motion machine, creating a vast sea of content that's bound to corrode any remaining sense of 'reality' not just on social media, but with other institutions, norms, and expectations.

Accelerating Change

So, this probably sounds...pessimistic.

I see it as inevitable. But that doesn't mean I look to the future with a sense of doom.

It's clear that we've entered an era where at least three forces will challenge existing culture and our political, economic and social systems:

• Attempts at climate resilience, and our discloation from place as a source of safety, community and abundance;

• A restructuring of how our economies work, driven in large part by the role of machines in replacing larger and larger swaths of human enterprise

• The radical restructuring of how, what and 'who' has the power and position to create. The shift in how creation happens will forever change our culture's structures for world-building (in media and entertainment), fact-based understanding (our ability to discern what is 'real'), work (what work is and who should perform it) and socially connection (who we interact with and where).  

The Mind-Blowing Strangeness of AI Creativity

Again, the restructuring of creativity will at first seem to be mostly 'performative'.

But there's something else.

And we're only very slowly starting to see its broad outlines: namely, that what the 'machines' create (regardless of the degree of human intervention, prompting, shaping, and manipulation) will often be very beautiful, strange, illuminating, perception-shattering, and...well...mind-blowing.

I've taken a very deep dive into generative AI, both so that I can understand the code, interfaces, limitations and controls; and so that I can try to grasp the 'product' of AI.

My observations aren't scientific. They're based on what I'm sure is faulty pattern recognition (based on decades of viewing and thinking about media, the Metaverse, human identity and creativity).

So it's better to think of these as intuitions, and thus open to discussion and revision:

• AI can be manipulated, filtered, controlled, prompted, layered, siloed, and trained. But regardless of how much you try to contain it, there is some aspect of AI that is its own thing.

• The creative output of AI often has the feeling of being hallucinatory or psychedelic. We'll try to train the models to be more "human-like", but I think that AI taps into a sort of creative Jungian super-id because it isn't a single creative mind - it's able to encompass a civilization-wide bank of content.

• The creative power of AI is placcid. Meaning, the longer you work with it, the deeper you go, the more unknowable it starts to feel.

• It comes up with ideas that, being just slightly adjacent to human creativity, start to shift the domain in which it's operating. Now - this one is hard to describe, and is again more of an intuition (and might be biased based on how deep I've gone). But you start to sense "norms" shifting, at least on a personal level. AI starts to challenge your perception of what an image 'should' contain, how an argument should be structured, or how a character should act in a one-man play. And you can't help wondering whether if they shift for me, they might not shift for others also.

Yes, working "with" AI is performative (I can see how I can accomplish certain tasks more quickly, and at scale). And it amplifies creativity because it's like being able to work with a talented art director, editor, researcher, designer and concept artist (all at once).

But AI also introduces a serendipity, beauty, bizareness, and even a psychedelia that is hard to describe, and shifts perceptions around how wide the range is for creative output.

When you can tap into a civilization-wide super id, strange things start to happen.

But finally, there's a membrane of sorts. It's the point at which you bump up against the unknowability of AI. And this feeling, this reality, becomes more pronounced the deeper you go.

Maybe people with more experience than I have 'get it'. Maybe it's like learning to code - you struggle, and then one day it makes sense, and your mental model has caught up with your level of skill.

But I don't think so. Because AI will never be in a fixed-state, and so we'll always play catch-up, we'll always be surprised by some new thing that the intelligence has learned, and the systems will continue to be unimaginably huge (beyond even their creator's ability to truly grasp).

AI As Media

Using AI to create is coupled with another strange intuition or observation. And to be honest, my thesis about this isn't fully formed.

The closest I've seen is what Sequoia Capital calls "personalized dreams" in its road map for generative AI:

It's the idea that entire movies or games will emerge from our behaviour, prompts and interactions. AI will create a streaming service (or game platform) with an audience of one: you.

But I'm not sure we need to wait until 2030 to get there.

Instagram and Discord are awash in "mini-dreams" already - whether pictures of synthetically created sci fi universes, fashion shoots, imaginary architecture, or very, very large breasted women (and, in fairness, extremely well-built men).

Today, it can produce an image or story. Tomorrow, entire movies or comic books.

But what if we expand our consideration of AI beyond the media it produces into thinking of AI as a media itself? What if AI is the next media evolution?

We shifted from printo to radio, television to streaming - but what if we frame AI as the next media platform? What if AI is what replaces video streaming, or eBooks, or Spotify?

AI Interfaces As Distribution Channels

When I generate an image with Midjourney or Stable Diffusion, the line between creative impulse, output and consumption evaporates to a very few minutes. The interface might be Discord, an app or a web browser, but the connection to AI within those spaces is itself a form of media.

Today, the product of that interaction (the photo or story) is often distributed through other channels.

I create an image using Midjourney (via Discord). I can then distribute that image to Instagram, or will be able to distribute my movie to YouTube. But the distribution into older channels is optional, and I have had no less of a media experience during the moment of creation.

And so I've started wondering how quickly we'll see the creation platforms themselves become the primary distribution channels. At what point does the optional distribution of output to the older channels start to disappear? Doesn't this make AI itself a form of media?

ChatGPT put forth a decent thesis (not that you can always trust the machine) that:

"AI can be considered a form of media as it mediates information and communication, extends human faculties, serves as a platform for interaction, functions as a 'cool' medium, and impacts culture and society. As AI continues to evolve, it is crucial to recognize its role as a medium and analyze its implications from a cultural theoretical perspective."

This view is one useful 'framing' of how AI will have a cultural impact, beyond the performative impacts of making certain types of work redundant (or their amplification).

And it comes with a tricky footnote, namely that:

AI Is The First 'New' Media To Swallow Up The Media That Came Before

AI is the first to circle back on all the media that came before it and restructure them, while at the same time achieving a new potential dominance as a media itself.

In other words, AI might achieve a reach that exceeds games, television, radio and print. If AI has reached 100 million users in only a few months, and if it starts to infilitrate more of the spaces in which creation happens and is amplified, then it is hypothetically on track to quickly become the dominant media platform.

But it will also infilitrate and reshape all of the media that came before, by changing how games, television, radio and print are produced in the first place.

This creates a vexing challenge.

Because it means that the distribution channels, already under siege (whether cable television or your local newspaper) will be corroded from within by this emerging distribution platform, which is itself able to generate new outputs for the old media.

Or, to put it another way, by way of example: first, social media will be corroded by AI-generated content, 'fakes', synthetic content, and artificial beings. And then it will evaporate entirely because we are able to interface directly with the dream-creating machines.

Sense-Making In The Age of AI

I resist the idea of letting ’AI is a media' become a dominant prism for thinking about what it all means.

Not because it won't transform today's media and distribution landscape beyond recognition (it will).

But because calling it a 'media' (or a performative 'tool' for that matter) minimizes how I think culture will change and react to the presence of the creative machines that now live among us.

Kurzweil famously predicted (in 1999!) the era in which we're living when he wrote The Age of Spiritual Machines. He said that:

"The salient issue is not whether a claimant to consciousness derived from a machine substrate rather than a human one can pass the threshold of our ability to reasonably deny its claim; the key issue is the spiritual implications of this transformation."

He was suggesting that the advent of advanced AI will transform human values, culture, and society. This transformation might lead to new philosophical and spiritual perspectives on our place in the world, our relationship with technology, and our understanding of what it means to be human.

I always took exception to Kurzweil's projection of a merging of man and machine (I still don't see it, and won't upload my brain if it becomes possible) but could never really argue with his projection of an exponential age.

But his broader point that human culture will be transformed, and perhaps at a spiritual level, feels relevant today, as we grapple with the implications of AI, how 'sentient' it might be, and the arrival of another 'creator' that is adjacent to human creativity.

Systems For Sense-Making

It seems to me that a lot of the sense-making around AI centers around a few profound lines of enquiry:

• How to make sense of the performative powers of AI, and what its efficiency means for our economies and capacity to do things

• How our concepts of machine-based learning align to our understanding of the human mind

• How to understand the implications of AI on our understanding of fact, its implications on our perceptions of reality, and how it may blur or distort the trust we have in institutions and media production

There are others, of course.

And yet even if we knew the answers to these areas of enquiry, they don't fully answer the question I touched on above: what happens when symbol-using behaviors emerge from AI systems into the world?

Lessons from the Metaverse

I have my own personal biases and experiences. As I think about how AI (and our creative uses of it) will play out, I naturally gravitate to my own interests and experiences.

And in particular, I feel like I have seen these patterns before, in particular with how sense-making has played out in virtual worlds:

• The Metaverse, or virtual worlds, even 15 years ago, tended to gravitate towards myth-making, story creation, and structural ritual-creation. This was in part because they create a reality which is no less real from being digital; and from their capacity to actualize human imagination and creativity, letting us easily explore ideas around identity, community and space/place.

• Layered with the affordances of virtual environments was the concept advanced by Tom Boellstorff, who argued that we were in the age of homo faber, rather than homo sapiens, or of "man the maker of artificial tools". He further observed that what was remarkable about this age was that the techne was embedded within techne...tools within tools.

In a loose way, AI extends the ideas from synthetic worlds: by adding artificial intelligences, it means that the ideas underpinning the early Metaverse nowunderpin everything.

Now, instead of something being no less real from being digital, reality itself will be no more real because it is physical (or at least, no more easily understood via the cultiural artefacts we use to understand it); our capacity to actualize human creativity is both amplified and joined by new creative intelligences; our concepts of identity, community and space/place will be radically challenged; and our role as homo faber is adjoined by a machine that can create back at us.

Just as we've seen in other digital spaces, and with the emergence of other tools and media, we will make attempts at sense-making. I am informed by (and perhaps biased because of) our experiences in the early Metaverse.

And yet I acknowledge that there's a difference, because this time we'll be trying to make sense of a creative force that is ultimately unknowable.

Mythos and the Age of the Spiritual Machines

Kurzweil was prescient in predicting the implications on human's self-understanding, and yet I still look for a framework or model against which to assess emerging forms of sense-making.

The closest I can come to is the idea that a new collection of mythos (mythoi) will emerge, be created, with their accompanying structures, rituals, symbols and cosmologies. They will be mythos that are direct responses to, and created in partnership with AI.

These AIMythos may become powerful cultural drivers, their mythologies will be equivalent or exceed today's storyworlds and story universes, they will inform cultural tribes, and perhaps propel what Ribbonfarm Studios calls 'theocratic capture':

"Theocratic capture occurs when an institution surrenders itself to a cult demanding unaccountable authority and agency on the basis of claims to privileged knowledge about the world, unaccompanied by demonstrations of the validity of that knowledge. A priest in the arena is a charismatic figure leading a theocratic capture campaign."

In other words, AIMythos won't always be benign, especially as people rally against change, or rally around people who have learned to harness the powers of the new machines.

Mythologies provide a discipline or framework for understanding both the mythos that is created because AI has arrived, and provide a powerful set of paradigms for the creation of new forms of story structures, paradigms for engagement, blueprints for game and world-building, and reference points for how cultural artefacts will propogate.

In this way, a mythos can be seen as a discrete set of observable events, outputs and responses, that can include:

1. Beings or characters, in particular the synthetic 'intelligences', or the heroes and creatures that are the product of those intelligences

2. Creation or origin stories and cosmologies, including the metaphors we use for how AI 'works', and the cosmologies that arise from the fictional or performative outputs of AI

3. Moral and ethical codes, often expressed in terms of AI safety or guardrails, but also in how human culture reacts, values and judges the outputs of AI

4. Rituals and ceremonies, including the rituals around the interfaces we have with an AI this is, in the end, unknowable.

5. Historical or cultural events, including the stories we tell about AI, and the histories that are created by AI (or by our use of AI as an amplifier for the human creation of new storyworlds)

The idea of an AIMythos, therefore, has a pragmatic purpose:

• It lets us tap into a socio/cultural discipline for the study of emerging responses to AI;  

• It helps us develop toolkits for the creation of new cultural ecosystems, in the same way that today's creators generate storyworlds and canons, and in the same way that fan art, fandoms or the creator economy are currently underpinned by the disciplines of world-building and storytelling

• It helps to resolve having a nomenclature that is sufficient to accompany the premise that we are now working with intelligences that are both creative forces, and ultimately unknowable.

On this last point, I think that an AIMythos needs to be delineated from modern ideas of myth-making or, at least, from mythologies that are purely human construction.

Hollywood may itself have a mythology, for example. And it's in the business of creating new ones, from Harry Potter to the Avengers.

But the AIMythos is more akin to pre-modern mythologies, because at it's core it will arise and be created by interactions with AI creators that have more similarities with an old tribal god, with its impetuous powers and whims, its need for offerings and tribute, than with today's writer's rooms or ad agencies.

AIMythos will draw more from our attempts to confront magic (understanding things beyond the realm of normal experience or understanding) than logic or science; and it will ultimately be our attempt to create archetypes, symbols and cosmologies during an age of spiritual machines.

The new AI are not gods, and yet they are ultimately unknowable, and they will exercise a creative and destructive force that will change our social and economic systems. And they will have, as Kurzweil projected, an impact on human spirituality - on our understanding of the powers that shape the universe and thus on our understanding of ourselves.

Change In Times Both Terrifying and Strange

The physical world will drive mass migrations and attempts at resilience; our work lives will often be shattered beyond recognition; and the ways in which we previously found meaning will disappear or be unreliable.

Media, tribe, rationality or human-created story will no longer be enough. When the old systems start to fail, the new AIMythos will attempt to ground us, to give us a frame of reference to understand the forces at play in the world around us.

The products of these intelligences will often be beautiful and strange. We will perform rituals, make offerings, attempt to codify our understanding of them, and gravitate towards some of their outputs because we're moved or entertained, enlightened or empowered, immersed or impacted.

As systems, some of these mythologies will appear, on their face, logical or rational.

You already get a hint of this in the idea of the 'prompt engineer'. They're the new priesthood and, in a similar way, they may have a codified skill set in talking to the new creators, but they're no more capable of a direct line to the 'mind of god' than your rabbi.

I don't mean to sound mystical. But there might be a pseudo-religious undertone to how some of culture relates to AI.

And there is some degree of magic, of mysticism and the psychedelia that can underpin divine experiences, and this derives, I think, because AIMythos will be our attempt at a kind sense-making that bears similarities to the pre-modern:  

• We will attempt to explain the unexplainable, the unknowable, through storytelling, lore, and metaphor.

• We will try to explain, codify or otherwise document our understanding of magic - the production of effects that are beyond the realm of normal experience or understanding

• We will try to find ways to relate to, engage with, control and understand intelligences that are, in the end, always at a remove, always beyond our capacity to know

• We will try to build new structures, societies and proto-cosmologies around power (especially power that we don't understand). At one time those powers were the forces of the natural universe, and we created gods to explain them. Today, that power is artificial and yet no less significant.

It will be an era of 'vibes', ritual, divinitation, ambiguity, emergence, and belief. It will be an age of wonder and a psychedellic-level of human experience.

What Is The Future That You Imagine?

What will AIMythos look like? At first glance, it might not look dissimilar to today's storyworlds, interactive games, or fandoms.

Imagine a music star

She has an origin story, a history, a set of talents. Her fans attend her performances, and how they respond has a partial influence on what tracks she records, the lyrics she writes, the style she uses. She's responsive to the market, and yet has her own creative voice.

Now imagine that this star is entirely virtual, created entirely by AI. She's driven entirely by models:

• One is trained on certain styles of music, and songs are created

• One is trained on her appearance, and photos and videos are generated

• One is based on her language, backstory, and 'canon', and she can interact with fans via an avatar in a virtual space, or through posts on social media

A DAO is created to shape what she does next. The 'votes' are translated into AI prompts, and the outputs are fed back into her models. Maybe she finds god, picks up a drug addiction, or becomes enamoured with modern jazz.

At one level, she's just a large-scale interactive character.

But what we've learned from AI is that there's something else: because the AI itself is part of the creative process; it will produce jarring moments or music that 'doesn't work', but it will also create moments of great serendipity, and will take unexpected creative leaps.

No matter how well structured the system is, how and what it will produce are unknowable.

Fans will create lore, stories, rituals and sub-cultures around her. But if we view all of this as simply an AI-assisted interactive adventure, or as a new form of story distribution, we overlook the fact that there is a level at which human agency stops.

The structures that built her, and the cultural artefacts, rituals and outcomes that are the result are better aligned if understood as a mythos (or, at the very least, a mythology within a larger AIMythos) than a new kind of interactive performance.

Imagine a Nike Sneaker

Nike used AI to create a new sneaker, the Scorpion.

Instagram is filled with images of imaginary Nike sneakers.

Now imagine a universe of media created around it.

We (perhaps sadly) ascribe human-created mythologies to what we put on our feet.  But what happens when those mythologies are created entirely by the machines?

Imagine that Nike really let loose. Instead of curating the output, they let it into the wild. The heroes, stories and rituals built around this sneaker aren't curated by humans, but solely managed by AI.

A feedback loop would be created between fans/consumers and the brand. The brand story would take some unlikely leaps, generating images and artefacts which no human might have intended (or attempted).

The cultural artefacts built around the sneaker would have more in kind with a mythos than a brand guideline or ad campaign, would end up with its own cosmology and sub-campaigns, and would create an unstoppable loop with consumers, endlessly evolving while remaining organic, like some sort of branded mycelium, spreading its tendrils through the media landscape.

Imagine A World

And now imagine a world. You don't visit it very often, but when you do you notice that it has changed.

The world itself is generative and might evolve based on how we interact with it; the creatures we meet might be 'intelligent', and their histories might be programmed to evolve; the logical underpinnings of this world will be even more opaque when we realize that it is a giant, organic, ever-changing machine, governed by rules we can only glimpse, and driven by value systems that we can only make ultimately futile attempts to decode.

It is strangely beautiful, but it's unlike previous human constructed game worlds.

It evolves on its own, through opaque connections to outside systems.

It isn't just a self-contained intelligent agent, it's an intelligent world. And how we respond, how we feel, the stories we tell about our adventures through it sound like the hallucinations made by machines.

There really no other way to describe the underpinnings of this world other than as driven by a mythos, and the stories we tell about it don't just challenge our capacity for storytelling or sense-making, they challenge the notion of what it means to be me.

Welcome To The Age of Mythos And Magic

And so I arrive at a working hypothesis for this new age:

• As we enter an era in which we interact with intelligences that we don't fully understand, human culture will become detached from the scaffolding of truth  that it had once depended upon, while (in parallel) being dislocated from the physical environments in which it had once been resilient;

• In a search for meaning and grounding, culture will start to be based on projecting archetypes onto the new intelligences, creating feedback loops that result in emergent, synthetic, but very real mythologies and proto-cosmologies;

• The Metaverse will be a powerful site for experiencing our relationship with these mythologies and intelligences. It will provide a civilization-wide coping mechanism for relating to and understanding those emerging intelligences and provide cognitive bonding with synthetic beings.

• However, the Metaverse will be less 'contained' than we once believed. Our experiences will seamlessly shift across devices, worlds, and interfaces.

• Therefore, the proto-cosmologies we believe in, and the intelligences we interact with, will be 'always on'. Our coupling with them will become a powerful cultural markers to others, and the lens through which we experience interactions and information, both in physical and digital spaces.

• As cultural markers, these intelligences, and the mythologies and cosmologies attached to them, will attract others. In other words, we will find our tribes. The intelligences will evolve in response to these communal movements.

• We will be relating to 'beings' that hold a quasi-religious unknowability and capacity to perform magic. This will naturally lead us to generate rituals, institutions, codification, interpretative methods, protectiveness, caste systems, and guild-like specialization.

• Simultaneously, other artifical intelligences will be performing 'productivity at scale'. Using their unstoppable creative and analytic powers, AI will transform almost every industry, from medicine to nanotech, manufacturing to content creation, farming to law. This will cause both unparalleled progress and massive dislocation. When combined with our attempts at climate resilience, it is not unlikely that entire political and economic systems will collapse.

• Productivity-at-scale might therefore create another reinforcement loop for the emerging mythologies, driving increasingly deep attention and attachment to the new god figures and cosmologies that are created by the symbiosis between man and synthetic intelligences, and perhaps replacing or supplementing the previous economic and political systems.

It's a strange new age, filled with both peril and promise. It will be unbelievably sad, and deeply beautiful. It will elevate human consciousness, as the bounds of our creativity are first amplified, and then expanded by our interaction with the super-id of the new machines.

And perhaps instead of our doom, the age of spiritual machines will let us re-examine what it means to be human, let us question the privilege we have held, and leave us more attuned to the worlds that we did not create, than the ones we created.

I didn't intend for this to be such a long piece. Thanks for sticking around (if you did).

During the writing of this post, I was informed by conversations with ChatGPT (both 3 and 4). There are very few direct lifts from that conversation, but you might find it interesting, or if you've never seen how it works, it might even be illuminating.

In particular, the chat generated some ideas about how AI will play out in the Metaverse, and created some useful markers for mythologies that I did not fully explore in this post.

You can find a transcript of that chat here.

This post marks the migration of my blog Out of Scope to Substack. I may continue to cross-post larger 'pillar posts' there, simply because I never fully trust having my stuff held by some other platform.

At one time, I wrote over 750,000 words about the Metaverse under the banner of Dusan Writer's Metaverse. They were strange, wonderous and often frustrating times. Maybe it's dangerous to brush off and attempt to excavate a past ‘vibe’, but I find myself at a similar moment of wonder and it feels appropriate. Thus the name of this publication.      

I really don’t promise to write very often. I have found that to be a bit of a treadmill that I fear getting on again.

When I do, it will be because I find it useful to put my thoughts into the wild, in the hopes of generating a conversation. Feel free to email me at doug@bureauofbrightideas.com or message me on Twitter.

Let's chat.