We’ve all seen creatures in clouds. Or Jesus on a biscuit. It's something we naturally do.

 

It’s called pareidolia. It's the tendency for our perception to impose interpretation on a nebulous stimulus. 

 

Our minds fill in blanks or add shapes to things as a means of creating understanding. When, for instance, we are shown Rorschach blots it's pareidolia being used in attempt to peek into our minds.

 

Experts believe pareidolia is an evolutionary survival mechanism. The sooner we can identify a threat the better. If we err on the side of caution, so be it. 

 

In nature we see various species exploiting pareidolia to their advantage. The false eye spots of the Polyphemus moth elegantly turn the tables on hungry and somewhat gullible predators.

 

Visual AI works this way as well. AI hallucinations are a form of pareidolia. As the machine works hard to interpret an image it begins to make best guesses at what it is seeing.

 

 It consults the data set it was trained on to connect the dot. If it were trained on a data set with a lot of dogs, images may contain thousands of tiny Dalmatians.

 

Daniel Ambrosi, an AI artist who you’ll hear from below, has become an expert on how computers process visuals. His awe-inspiring artworks give us a clear glimpse of what is going on the the mind of AI. His

images are striking, and, yes, hallucinatory.

 

We are sense-making machines. But it is a two-way street. Not only do we have an instinct to quickly identify things that may be fast-approaching threats, but we also want to impose our own personal outbound sense on others. A sort of inverse pareidolia is the order of the day.

 

Through the use our phone's glamorous photo filters, or by using AI to strategically and seamlessly edit our pictures, like the Polyphemus moth we are generating false views of the world we hope others will quickly accept as reality.

 

We are entering an age of illusion that is fast revealing what deep down we always knew. That what we call reality may be nothing more than a well-curated hallucination.

 

Welcome to Nextness. More than a magazine, a mindset.

The dream goes something like this. In the not-too-distant future we can create video content instantly. Trained on our favorite movies, we can write prompts that can seamlessly insert ourselves into the storyline of our favorite shows. Then we can watch and share programming starring, yes, you and I.

 

It's sound so sci-fi, but it is here – it has already happened. 

 

Brought to us by the folks at Fable Studio, the unauthorized South Park episode dubbed Westland Chronicles is a surprisingly entertaining episode of the prime time hit.

 

For the people at Fable who wrote themselves into the episode it is even more satisfying. 

 

This is not a stunt. What Fable is trying to solve for is a system that will instantly generate new content and storylines. They call this Showrunner AI, where the “user" becomes a kind of omnipotent storyteller.

 

The Fable team, who recently won a Peabody Award, have approached this as a media science experiment and have published a research paper – To Infinity and Beyond: SHOW-1 and Showrunner Agents in Multi-Agent Simulations – outlining the potential of their “Showrunner AI” to write, produce, direct, cast, edit, voice and generate episodes of television. 

.

For this system to make truly entertaining content it must overcome numerous challenges.

 

The first is called the Slot Machine Effect:

 

 “…where the generation of AI-produced content feels more like a random game of chance rather than a deliberate creative process.”

 

When AI simply spits out the same stuff in different combinations things become stale:

 

“Our attempt to generate episodes through prompt-chaining is due to the fact that story generation is a highly discontinuous task. These are tasks where the content generation cannot be done in a gradual or continuous way, but instead requires a certain ‘Eureka’ idea that accounts for a discontinuous leap in the progress towards the solution of the task. The content generation involves discovering or inventing a new way of looking at or framing the problem, that enables the generation of the rest of the content.”

Another challenge is the 10,000 Bowls of Oatmeal Problem. If AI creates things without human oversight it will, overtime, create an astonishing amount of content that seems the same. We are seeing this now.

 

Yet what many consider a flaw within AI can actually help solve this problem. AI “hallucinations” can be additive to the plotline, introducing unexpected twists and turns that can disrupt the tendency of AI towards predictability:

 

"Positive hallucinations" or happy accidents of a complex generative system can be a good thing. Surprising the user by balancing and changing the phases of certainty vs. uncertainty helps to increase their overall engagement.”

 

As ever, creative intentionality will be key.

 

When we are involved in stories, we are continually searching for patterns and making predictions. When we guess what is coming down the road, so long as it’s not too obvious, our brains get slammed with a big hit of dopamine.

 

Part of this game revolves around guessing the intent of the storyteller and trying to predict what the moral of the story might be:

 

“Another factor that keeps audiences engaged while watching a show and what makes episodes unique is intentionality from the authors. A satirical moral premise twisted social commentary, recent world events or cameos by celebrities are major elements for South Park. Other show types, for example sitcoms, usually progress mainly through changes in relationship (some of which are never fulfilled), keeping the audience hooked despite following the same format and formula.

Intentionality from the user to generate a high-quality episode is another area of internal research. Even users without a background in dramatic writing should be able to come up with stories, themes, or major dramatic questions they want to see played out within the simulation."

In addition, the Fable team believes the Showrunner system could guide the user by sharing its own creative ideas, or prompting the user by asking the right questions. 

What Showrunner AI demonstrates is that AI can be a massively powerful collaborator in the making of content. 

 

Our world will be very different very soon.

 

The obstacles to making instant content are hardly insurmountable and they are being solved at this moment. 

 

Machines that can effectively tell stories with scant human input will be the most disruptive force in electronic media since television.

 

Whether it’s writing yourself into existing shows or movies, or creating your own original content, AI tools like Showrunner will change everything.

 

Check out the South Park Westland Chronicles here:

 

And read the research paper here: https://fablestudio.github.io/showrunner-agents/

There is a well-known pose you see in art galleries. It’s the contemplative stance – feet firmly planted, head slightly askew, one hand attached to face, thumb cradling chin, index finger forming a question mark just below nose.

 

This is how we view art.

 

The work of Daniel Ambrosi completely shatters this posture. To look at one of his Dreamscapes becomes a dance. People move forward and back, then in again and sideways, always squinting then standing aback.

After watching people at his recent solo art show for 7 months, it came down to a four-step process:

 

“Step one. Wow. They all start with the word – Wow. Then they look for a few seconds, they tilt their head, and they go, ‘What is that? Is it a photograph? Is it a painting?’ They walk up to it, and they look at it from the side they go, ‘Is it 3D? Is it moving?’ The second step is this curiosity like ‘What am I looking at?’ The third step I get are the superlatives: ‘Oh, are you the artist? It's magnificent.’ And then the final step, which is super consistent, I hear this all the time is, ‘I've never seen anything like this!’”

 

What Daniel creates are digital dreams, enabled by a fusion of computational photography techniques and choreographed computer hallucination.

 

Daniel’s goal has always been to capture the transcendence of nature on a canvas of light. Surrounded by the cathedral-like walls of the painted desert he would take dozens images and blend them together.

 

“As an avid hiker and traveler and skier, I seem to be especially sensitive to these things, these visions that happen in certain places and certain compositions certain times of day, that it just hits me like a hammer and besides being blown away by what I'm seeing, at that point, I just feel compelled to share it with others.”

He wanted to bottle the feeling of awe he felt in the world’s most sublime places. He succeeded. But, for him, there was always that nagging missing piece. The intangible feeling of unlimited wonder sometimes remained out there in nature. How could he bring that back?

 

“I was getting part of the way there in terms of reaching people and getting that visceral reaction of taking their breath away. But it occurred to me when DeepDream came along that this might be a way I can tap into a bit of that curiosity as well.”

 

DeepDream is a process invented by Alexander Mordvinstev and Mike Tyka at Google in 2015. Author Arthur I. Miller describes DeepDream as, “It allows you to see a vision of the world through the eyes of the machine.”

 

When the machine is in the process of trying to figure out what it is looking at it conjures some very trippy stuff. At certain stages images take on a Starry Night quality and become highly impressionistic.

 

The key word is trippy:

 

“And it's not unlike in a hallucination, because it's trying to make sense, you know, out of what it's seeing, and your mind is kind of constructing things out of random until they stick, right?”

 

As much as we might associate hallucinations with recreational drug use, or CIA experiments gone wrong, psychedelic visions of this variety arose as a survival mechanism:

 

“It’s called pareidolia, which is a trait that I think probably a lot of animals have, it's, from what I understand, an evolutionary survival thing. You want to imagine things, especially things that might hurt you. So it's better that you imagine you see a tiger in the grass and be wrong than not seeing the tiger in the grass.”

 

Daniel’s art is conjuring the layer in DeepDream that best captures the essence of what he saw and felt in the field. These are artistic choices that are guided by a preternatural connection with what is really out there:

 

“I'm a metaphysical idealist. And I think, you know, rather than consciousness emerging from matter, I personally believe matter emerges from consciousness. And while there indeed is a real world out there. I think it probably is nothing like what you think it is.”

 

Daniel’s work makes us rethink how we see the world. And how machines may view it. His work forces us to reconcile two ways of seeing things. Its why people step forward and back deeply apprehending both worlds until a synthesis of the two forms.  In that moment, they see a new blended world that until then had only existed in the mind of the artist.

 

This is where the magic lives:

 

“Because that's like, the hardest thing, you know, especially for an artist because you feel it, you see it in all dimensions, and you just hope that you can transport that experience to somebody else.”

As much as we might associate hallucinations with recreational drug use or CIA experiments gone wrong, psychedelic visions of this variety arose as a survival mechanism:

 

“It’s called pareidolia, which is a trait that I think probably a lot of animals have, it's, from what I understand, an evolutionary survival thing. You want to imagine things, especially things that might hurt you. So it's better that you imagine you see a tiger in the grass and be wrong than not seeing the tiger in the grass.”

 

Daniel’s art is conjuring the layer in DeepDream that best captures the essence of what he saw and felt in the field. These are artistic choices that are guided by a preternatural connection with what is really out there:

 

“I'm a metaphysical idealist. And I think, you know, rather than consciousness emerging from matter, I personally believe matter emerges from consciousness. And while there indeed is a real world out there. I think it probably is nothing like what you think it is.”

 

Daniel’s work makes us rethink how we see the world. And how machines may view it. His work forces us to reconcile two ways of seeing things. Its why people step forward and back deeply apprehending both worlds until a synthesis of the two forms.  In that moment, they see a new blended world that until then had only existed in the mind of the artist.

 

This is where the magic lives:

 

“Because that's like, the hardest thing, you know, especially for an artist because you feel it, you see it in all dimensions, and you just hope that you can transport that experience to somebody else.”

 

 

 

 

In his revealing book "How to Change Your Mind" Michael Pollan writes, “Carl Jung once wrote that it is not the young but people in middle age who need to have an ”experience of numinous" to help them negotiate the second half of life."

 

This is the journey Michael Pollan went on and wrote about in “How to Change Your Mind.”

 

In this age where we hear so much about AI "hallucinations" we begin to assume it is a bad thing. Strangely enough, we find it may be AI's most human attribute.

 

Before Michael went on his first LSD trip his psychologist explained, “Kids perception are not mediated by expectations and conventions in the been-there, done-that way that adult perception is; as adults, she explained, our minds don't simply take in the world as it is so much as they make educated guesses about it. Relying on these guesses, which are based on past experience, saves the mind time and energy, as when, say, it's trying to figure out what a fractal pattern of green dots in its visual field might be. (The leaves on a tree probably)”

 

As we've seen from the work of Daniel Ambrosi, we perceive the world very much the way AI does. This understanding will most likely provide new into what goes on in our own heads.

 

Spending a little more time understanding the mind of AI just might help us regain a small bit of the child-like wonder we once possessed.

 

Sponser

AI changes everything. Including storytelling. Nextness Volume 6 is brought to you by Storymachine. As a leader in AI content, Storymachine scripts, films and delivers everything from branded content and commercials to corporate masterclasses and training films. If you are looking to unleash a new kind of storytelling, Storymachine just might be your jam.  storymachinefilms.com