Retrieve The Intangible & Gen AI

This moment is one you can split with a clap.
And when it splits you end up with many pieces.

If you looked at the lazy Sunday afternoon Florentine crowds walking by, you would never guess the informational turbulence that was crossing the world in this period. It was February 2025, a time that will make history as a peak play of populist politics, in which everything flipped front to back. I was sitting in a café writing. In front of me I could see a mix of locals and tourists from around the world that had not caught up with the sense of division being sown back there, in the crypto power grabbing elite space.

In the backdrop to the barrage of information and land grabs there had been a similar aggressive application of technology. Working their way around limitations of copyright, systems had been devised to effectively mock the style of original works of any number of known writers, illustrators, photographers, designers… Creative professionals of all kinds, living and past could be invoked through a prompt, recreating a synthetic version of their output.

While initially these were cute toys, their performance then improved and spread out in versatility leading to speculation that many creative professionals would have a difficult time with their career ahead – many people I knew had already lost their jobs.

My cafe writing was following a mission to try to refocus and give a direction to my artist friends. Beyond politics, I sensed a difficult world ahead, in which creatives would need to dig in to understand where they could bring their value.

Just then, my musician friend Agustin sent me a message from his South American tour: “Can you send us that list? I think we’re ready.”

Erika and Agustin, called Sinedades, were visiting South America planning a documentary. Their Latino-Samba-Poetic blend of music had grown up in Tuscany, Italy while looking across the Atlantic to musicians in Argentina, Brazil… Now they wanted to tell the story of these influences with an eye from the outside.

In any case, I appreciated their knack for poetic scenes and so I had asked for their help in creating some short performances. I would write a list of ideas they could interpret and create some short video snippets that could not be easily generated synthetically. I imagined them being messages that an AI would not understand but that would be intuitive for humans. Content that would typify what you could not synthesise, and that would give a direction for future aesthetics. My purpose was to try to look ahead at how culture might change in response to these tools.

I was following the intuition that the easier it was to create new output: images, videos, animations, music, with a simple text description, the less the value of the creation. And given this – my next question was: how would culture adjust? What sort of expression would help artists retain their exclusive value compared to the morass of synthetic output?

Follow the money, they say… and the world online was not much different than it had been. Advertising that targeted what people paid attention to, was still the core, but many things would have to change dramatically. Culture and attention is fickle: it would drift toward authentic, exclusive, hard to reproduce content. You could see that whenever a new video effect, or technological gimmick would appear, it would not last long before receiving a sort of backlash reset movement.

Humans looked for dreams of things that were singular moments of existence, not repeatable mechanically, not generated through a text prompt. They looked for stories that underlie the work that allow us to immerse ourselves in the piece we are looking at. We like to know more about who made the piece and why for example: any additional context matters.

I typed back some excuses: “Yes, Agu… I am working on it… nearly done. The last draft didn’t seem right, so I am reworking it. Anything you can do will be super appreciated.”

I imagined Agustin in a different world, where technology was not as pervasive in mediating connections. I saw videos of beaches in Rio: a human liquid chaos of people living, cooking, eating, sleeping, working under temporary structures of umbrellas that covered the beach completely without a break at any point. Personally, it felt a bit of a nightmare, but on the other hand this was so different from my WhatsApp/Messenger/Email/Slack/Insta madness… I could receive a message in any of around twenty ways, and they were all basically – just messages. Somehow, I doubted the beach goers in Rio had much more than WhatsApp, and I was jealous of this simplicity.

I was remembering the magazines I grew up with in the ’70s and ’80s. They were thick and full of gorgeous places and models, it all looked so perfect and so aligned. One could go to dinner with friends and know that the reference for Luxury would be a shared idea of the perfect vacation. An empty beach, white sands, the perfect mountain, a gorgeous village with lovely cheese and wine.

All of this would change because our AI technologies completely absorbed our past production and could now spew infinite endless versions of perfect photos of perfect vacations and smiles at an unhuman scale. We would quickly become completely bored with this and seek something scarce and deeper.

And so, it felt a bit like the world was on the cusp of committing a sort of creative cultural suicide because no matter what AI created, the next generations would find it boring and worthless due to its oversupply.

A bit later, I got a message back: “No problem, we’re here… Going back to Brazil in two days. Argentina was incredible.”

Families and couples were walking by in Florence. It was low season so tourists were only a small fraction of the people I could see out Sunday afternoon. And it was cold February which meant lots of coats and pink cheeks.

The crowd seemed completely disconnected to what I was imagining. Something about living in a medium sized city in Italy, created a distant technological bridge to the rest of the busy AI development chatter I was getting. There were posters up fighting back against the Algorithm, from student organisations, but I could still run into people that worked as designers, illustrators, photographers. Something I had seen disappear almost completely overnight in the advertising circles in London. When I asked these creative “artisans” where their funds were coming from, almost inevitably they were connected to either the food business or weddings, particular local exclusives.

Somehow, I didn’t think synthetic production would affect wedding photography that much, so maybe they were safe.

So, then there was the problem to try to explain and articulate what would be left of value in a world flooded with this synthetic production. What room would AI leave humanity to play with? And how sure could we be that we still had meaningful free expression as AI grew more powerful? Would it feel like being awash with a waste of disconnected memetic jumbles, or a special new world of ideas?

I called the idea that humans still maintain some edge on synthetic creativity The Intangible. The term represented what would be impossible to capture and recreate with synthetic creativity, which is another term I preferred over AI because it gave me a clearer picture of how creative processes were reshuffling past human production to create something new but losing all the context of their originals in the process.

I got distracted from my writing and dreamt of my friends Agustin and Erika in Argentina. Agustin had been born in Argentina but moved to Italy when seven years old. He and Erika had been playing music since they were teenagers. They were a bit like a connected whole person with a knack for delivering poetic moments, a sort of performance magic.

I had to ask: “What happened in Argentina?”

I am not sure to what extent they even understood how their magic worked but I’d experienced it in so many situations that I knew it was not chance. And it was not even just a matter of musical performance. There were other dimensions to what they did that would strike deep into their audience.

Agustin replied right away: “We played in a village square. I told them my story, that I had come back home. We played Rosas and everyone was crying…”

I replied: “Wow…”, but a tear dropped from my eyes imagining the scene and remembering the song.

I resorted to something approximate, to try to organize this list of scenes. I would try to do the best I could with a list of ideas of what makes us human, however incomplete. I think whatever sketches I might create they would fall under one three areas:

A) Intrinsic human qualities tied to consciousness for example:

We are independent entities; with a sense of presence we call consciousness.
We value experience itself as a possible reward for curiosity.
We value pleasurable comfort and beauty.

For this category a sketch could involve finding a connection to the audience that is viewing through simple sensory experiences. An example might be to create and experience Silence. I could see this in a space that has a lot of echo and reverb and the camera shows images and subtle sounds of someone walking up the stairs, then their expression and the subtle sounds of their breathing. What matters is that when we look at someone that is silent they are never completely silent, and we imagine what is happening inside their head.

Other sketches might involve different states of consciousness like sleeping, looking out at the ocean waves mesmerised, or any number of other curious behaviours that someone might take on as experiments. Things like playing with pebbles, paying intense attention to scratching an itch, watching ants move about or standing on a chair for ten minutes. But even simply someone experiencing beauty.

B) The beauty of imperfection so:

We like to realise how our senses can fail and leave us with a sense of magic such as optical illusions, or other sensory illusions
We like patterns and messages we don’t understand (poetry, music)
We enjoy ambiguity, it makes us laugh and cry

For this category a sketch should play with perception by distorting what initially might seem to be a message, but that then becomes something different and in itself a new message of sorts. An example might be a sentence, or a word that is repeated so that where initially it has one meaning it then turns into another meaning and has a whole new meaning overall.

Rhythmic patterns and their meaning also could be a way. Imagine you asked people to answer questions with rhythms, forcing them to show happiness, sadness, patience, tiredness with a beat. This reminds me of a psychological experiment where they asked people to choose the right name to two drawings: one was made of hard spiky angles, one was of rounded circles. Many more than expected named the spiky one “kiki” and the round one “bouba” showing there is an innate congruence between shape and sound: Synesthesia. And the Bouba/Kiki effect seemed to go beyond cultural differences.

C) The magic of human contact and:

We value social interaction because we like feeling part of something
We enjoy stories that have a mysterious echo of generality
We like to create and play games just to understand how the games work
We treasure originality and authenticity because it shows we are conscious

For this category I could use hand gestures, different kinds of non verbal communication for example. Imagine a dialogue that is going on between two people just with their eyes. Or imagine two people exchanging sentences with different gestures. Or imagine a game where each side has to respond to the previous gesture with another invented one. The point being that there is more than just verbal data exchanged between people, a whole world of non verbal aspects are also exchanged. In some ways exchanging deeper context we enter in a kind of shared consciousness with others, a state in which we are present in a similar way as our friends.

I ended my messages Sunday night late: “Agu, sorry I am so late in sending this. I ended up writing a sort of fictionalized narrative that I am putting on my Blog and will link to in LinkedIn hoping to get some weird commission to create AI proof content guide…”

Then: “So I will paste the whole story but it’s full of stuff you don’t need to worry about. Read it if you are curious and tell me what you think. Instead, I list the clips I would love to see you figure out how to interpret here as brief as I could and I give you total freedom to play and interpret.”

So I wrote: “1) Silence and Beauty: particularly I am interested in silence that isn’t perfectly silent, more what we think of as silent. I think it can work well with the documentary too. I keep imagining someone tip toeing barefoot up the stairs and we can barely hear these steps. If you can do a few different clips, would be great.

An AI would not know how to make logical sense of this silence. And even if it could understand the general pretext that this is serving a sort of spiritual experience it would now have a sense of what the conscious experience of silence is like.

2) Synaesthesia Rhythm: using rhythmic patterns of words, instruments or percussions to express things that are not the medium itself. Like what is the beat of sadness, or what does it feel like when you hear a word repeated in different ways, with different tones. Look at color, images, drawings, emotions, light, movement of the body, anything that creates a new broader context that is hard to fully capture in words.

Please note that in this section we should not emphasize communication between people but rather the audience is the viewer which gets the sense they are understanding something from patterns, sensory tricks, mix of mediums. It should be a bit like watching an optical illusion.

An AI could not fully understand this because it has different sensory systems. It would not perceive the same optical illusions we do.

3) Communication and Poetry: capture gestures and expressions to convey different meaning. The idea here is to make the viewer realize how rich communication can be and how much is shared that goes beyond words.

An AI would not know how to balance the merging of all the aspects of non-verbal communication with the verbal to get a sense of what is going on. In many ways this relates to 2 because we get a lot more out of communication from shared sensory systems. I can gesture a caress and the other will easily imagine what it feels like.”

Finally: “Thank YOU!”

Value Subverts Itself

Screen Shot 2017-05-25 at 23.04.15.png

The way people see
technical constructs and systems,
as permanent fixtures of value,
to invest in,
safer than countries, land or business,
is exposed to the same disruption,
that revolutionised the original systems
in the first place.
Why would you see more impermanence,
in an​ encrypted special number,
that someone stores for you,
somewhere safe and trusted,
whilst accelerating technologies and tricks,
try their hands at guessing.
Quantum computer waits,
to tell you the tulips market,
is gone.

A common technologist philosophy is one that tools don’t really need people to run them. In particular, I find myself in a ​frequent argument with friends about speculations on the market of encrypted scarce members of a blockchain. These friends, talk like the buffaloes, the stories are about the imagined possibility of higher and higher value, but they do not recognise that at some point, there is a cliff, and little chance to turn around. We would rather hold onto the dream of what is likely to be dust than cash in, and get something. This problem appears to be intrinsic to human behaviour, we seek collective tragedy somehow. The story of collective tragedy is perhaps more of a heroes journey than being careful and boring.

All good, but what kind of intelligence is this?
Will the superintelligent gamble the same way?
Is it risk taking that drives new evolutionary candidates?
If so then the superintelligent must be failure machines…

Digital Phenotyping

Screen Shot 2017-05-22 at 23.45.50.png

The language of science,
can transform,
like a superhero,
with a costume.
Like Batman out,
from the cave,
springing onto the city,
ready to catch some,
troublesome guys,
that foul language,
with their imprecise,
definitions and every day,
senselessness.
But underneath the costume,
sometimes you'll find,
similar-senselessness,
of purpose,
and weird looking socks.

When you see fancy words it is usually worth trying to make sense of them. As I first read the word “phenotype”: something like a measurable trait that comes from genes interacting with the environment. And then “digital phenotype”: some measurable traits that are recorded digitally coming from genes interacting with the environment. But in some cases, you will find, “digital phenotype” is more like measuring something about who you are from an analysis of multiple streams of data: where you go, how fast, when, etc… So the word “digital phenotype” could mean “measuring everything of a person to decide something about them”? Maybe with a subtle focus on their health, or identifying key patterns that might relate to their gene expression?

The problems with this use of language are really quite frustrating:

  • The Gene as a bit of code that defines you is too simple
  • The Gene interacting with Environment appearing as patterns of your daily routine is a bit far fetched and also too simple.
  • The measurement of daily data sources, and “Big Data” analysis of where you walk, to identify “Digital Phenotypes” seems improper: could be junk, could be unrelated to Genotype.

What are we saying with these twists of language? Could it be we are trying to convince people there is power and opportunity to be gleaned from evaluating everyday routines of people? If so, I am not quite sure this is the spirit of scientific enquiry.

Carpet Dreams…

Screen Shot 2017-05-18 at 11.05.29.png

 

The cold floor,
the traces of gum, and dirt,
from old boots,
and fancy sneakers,
the ash of a cigarette,
a bit of forgotten cereal flake,
the clawing of a cat,
the piece of a spaghetti,
some drool from a child,
all are not so much,
of a big deal.
If one can learn,
and try,
to find,
a way to fly.

Maybe it is the opportunity to learn that most values being human. But in many ways implicit in the learning is the potential to apply this knowledge to solving problems ​and improving practical circumstances. Does this mean that technology drives the need to learn? If so, it would be our own tools that potentially divide us by demanding hyper-specialisations? Or what about learning that leads you to a more balanced perspective on the relative benefits of different technologies and how they transform humanity?

Sir, General AI’s Cryptocurrency

Screen Shot 2017-05-03 at 00.17.14.png

The idea of secret number codes,
representing unique,
special moments,
of your work life,
you can trade,
for exchange,
of shared effort,
to create a better world,
of trust between code sharers,
no matter their identity,
greed, location, or creed,
or physical presence, indeed,
sounds pretty and nice,
and something,
the general AI,
will mostly,
approve of,
for now.

It is very difficult to evaluate the potential negative risks of a badly implemented new currency system, in particular, if you consider how it might relate to the impact of algorithms that decide its course. The problem is not only one of values but one of global stability, in that it is possible that the side effect of a euphoric adoption of a new currency could lead to decrease in value in others, which might in hand create a situation of exclusion to those that rely on conventional currencies. And on the other hand, the new currency users might find themselves in a situation of technological lock-in.

And if we want to caution on the risk emerging from General AI, we might at least try to design a new currency around a currency intrinsic to humans. What if, for example, you could trade your human attention in return for currency?