Welcome to Emerging Futures -- Volume 138! AI, Algorithmic Mimicry and More-Than-Human Creativity...
Good morning abstract processual assemblages,
This week, we have to start by saying, It's been quite a week. There’s been a little too much going on –
...and so we come to Thursday and the evening a bit further on the side of exhaustion than we would have hoped.
But let’s put that aside for a moment.
Last weekend we did get out to Printed Matter’s NY Art Book Fair – it is always quite an interesting spectacle. They close off the block, and there is a great street party. Music, food, events, and most enjoyably, as all sorts of people just show up with a few interesting and often very odd books, put them on a sheet and start selling. It makes for great conversations and the occasional really interesting find.
Inside the fair are four stories and a roof deck of far too many people and far too many tables from small presses and individual bookmakers/authors. Each table being a world in itself with a unique perspectives and ethos. There is only so much of this we could engage with before things in the body just melt. Then it was time to sit by the Hudson River and let things drift in the afternoon sun.
In the thick of things we saw some old friends whose work is of interest as we explore questions of AI and creativity in this newsletter:
Chang Yuchen - We have used their Book Book as the basis of some workshops we do where we make notebooks as part of the creative practice. Their work goes far beyond this into experimenting with form and meaning.
Audra Wolowiec (and Gravel Projects) – Audra is exploring the spaces and practices that exist in parallel with sound. We picked up the “book” SOUND that explores the extended ways we visualize sound.
Laurence Kumpf and Blank Forms – Blank Forms is one of our favorite places for experimental music in the city. If you are in the city or environs – mark May 7th on your calendars as they have a very interesting event on India’s first electronic music studio (1969) that will focus on ”The collection of music, tape experiments, and field recordings that encapsulates a utopian period as India explored a new post-colonial sonic imaginary within an emerging progressive vision for art,design, and technology. This was a moment in which the dialogues between Eastern and Western modernisms converged, and the pioneering spirit of the NID developed a unique cultural exchange with the New York avant-garde.”
We also made some great new discoveries; we picked up some wonderful, moving, and timely works by : Leila Abdelrazaq and Ellie Irons.
The reason to mention all of this is not just to call out some really interesting work in relation to creative practices in general. But this week, as we begin to explore questions of AI and creativity, it is very helpful to bring with us some touch points to think beside.
Much of our life is profoundly ambiguous, lacks any clear symbolic content, and requires us not to “interpret” but to actively co-create the possibility of a possible form of co-invented meaningfulness. This is not something a computational algorithm can do. Engaging with works like Chang Yuchen’s Non-Time Calendar—a very simple object, just one sheet of paper with few words and twelve simple drawings—but one of great ambiguity—can be very helpful.
The space of “Artificial Intelligence” is both quite big and developing quite rapidly. This week, prior to diving in to explore things over the coming weeks, we just want to lay out some basic general thoughts on how we will proceed:
ONE: We are looking at things from the perspective of creativity. Our interest is not to deliver a verdict on AI from some general God’s eye perspective; this, as a perspective in general, is absurd, and even more so when engaging with such a dynamic and emerging set of technologies. Our curiosity is in how AI can engage with questions of creativity—the processes by which the new comes into being.
TWO: It is not about humans vs. machines in regards to who is authentically creative. There is a lot of discussion out there that claims to pass judgment of the kind: “only humans are authentically creative – AI can never do this”.
We recognize that creativity is not some internal individual human capacity and that creativity is a worldly phenomenon. We find creativity everywhere in the universe; it is an ongoing processual reality. We participate in it, but it certainly is not ours. Thus, claiming humans are creative while something else is not—this is a non-starter for us.
Setting up Manichean dualities, where we divide reality in two, code one as good and the other bad, and then put one in opposition to the other, is almost never a helpful starting point to experiment with things.
All life—all reality—is processual. And as such, there are general principles and general patterns to many of these processes that can be found at varying scales. The repetition of such patterns means that they are loosely “algorithmic” (but not in the computational sense of the term). One could loosely consider creative processes that we find in biological evolution, for example, to be algorithmic. Darwin’s mechanism for creative evolution (variation + selection) would be an example of this. And as such, creativity and algorithmic processes are quite connected. Gary Tomlison in The Machines of Evolution and the Scope of Meaning explores this in very helpful ways. He develops Deleuze’s approach to this form of algorithmic reality with the concept “Abstract Machines” (“by which I mean something so simple as to be almost ineffable: conditions that, if met, set in motion a process”). He argues in this book that meaning in life emerges via four nested abstract machines: Natural Selection, Niche Construction, Mediation, and Semiosis. Creativity in short, as types of processes that are found throughout reality operating at various scales, do lend themselves to being engaged with as “algorithms” – which is to say in “machinic” ways.
As humans – as living beings we are always “more-than” forms of beings. This is precisely what 4EA cognition means. We are necessarily embodied, enactive, embedded, and extended beings (the four “E”’s). To live, we are hybrids of many tools (extensions) and environments (embedded). There is no way to be alive that is not such an integral hybrid. Computational systems are one of these extensions; they are a form of augmentation.
THREE: Innovation and more-than-human creative systems have a long history. As follows from the previous point: Human creative practices have long utilized deliberate algorithmic and computational systems to great effect. The examples that we have explored in the past have been the work of John Cage, and Georges Perec. A great book on this topic is Margorie Perloff’s Radical Artifice. (You can also query our site to see all of our discussions about their work – we have a great search feature embedded on our website).
It is also possible to apply the generative logic of creative evolutionary algorithms to non-computational forms of machinic creativity to great effect. And to do this in ways that Large Language Models such as Chat GPT cannot do. We have written about this in two previous newsletter:
FOUR: We need to define “intelligence”. “Artificial intelligence,” as understood by the champions of forms of machine learning that we see in large language models (LLM’s) has nothing to do with intelligence as understood in living beings. Artificial Intelligence is better termed Algorithmic Mimicry.
The project of developing an Artificial General Intelligence (AGI) is one that is certainly possible. But the current claims for LLM’s being at the cusp of realizing this rest upon profound misunderstandings of what intelligence is and how meaning is made. Here it would be useful to quote Johannes Jaeger and his excellent (and highly recommended) article, Artificial Intelligence is Algorithmic Mimicry at length:
“First, organisms are self- manufacturing (autopoietic) physical systems, which can identify, set, and pursue their own intrinsic goals, while algo- rithms are purely symbolic machines that are dependent on a suitable computational environment and target functions that must be provided by some external agent. Second, the prevalent computational architecture of today maximizes the isolation of software from hardware, meaning that interactions with the physical world require externally provided sensors and effectors, while no software-hardware distinction exists in organisms, which are embodied in a way that enables more immediate exchange with their physical surroundings. Finally, algorithms exist in a small world, in which all possible problems are well-defined, whereas organisms live in a large world, where most problems are ill-defined (and some are probably not properly definable). All of this goes to show just how different living systems are from algorithms.
This means that algorithms and living beings have very different capabilities and limitations. Surprisingly, this fact is often overlooked in discussions that compare algorithmic mimicry with natural general intelligence. It is true that algorithms outperform humans in many tasks. Those tasks are typically well-defined but hard for us to solve, especially if they involve large amounts of calculation, require substantial working memory, and/or involve high-dimensional search spaces. Familiar examples are strategic games like chess or Go, complicated scheduling or planning tasks, intricate mathematical proofs that require a large number of steps, or (in the case of LLMs) filling in the blanks of a text based on massive amounts of correlations between words and/or phrases in a training dataset. Because humans are cognitively rather limited in these areas, never having evolved the ability to solve complicated problems of this kind, we are easily impressed by the algorithms’ performance. Sometimes, this causes us to lose perspective, to attribute capabilities to these machines that they cannot possibly have because of their architectural limitations.”
That said, understanding machine learning to be something other than intelligence does not mean that it does not have very interesting roles to play in human practices of creativity. Nor is it to dismiss the project of an artificial general intelligence out of hand. There is no reason to think that other approaches to artificial life – perhaps more enactive ones, will not lead to some very unique form of AGI.
FIVE: It is never about AI in isolation – nor is it about the search for the correct understanding of “what is AI”. Nothing is reducible to what it is. The key questions are always: “what can it do?” “What is it doing?” and “What else can it do?”. These are open questions and experimental ones.
Critique and dismissal are two very different practices. There are many armchair prognosticators out there that wish to dismiss AI because of what it is and especially because of what it is not. But to understand that AI has nothing to do with intelligence and rather is a form of Algorithmic Mimicry is not to dismiss it – especially in relation to creativity.
The meaning of anything is the emergent outcome of the relational dynamics that it is part of. All of the various forms of AI are no different. We need to be both cautious and actively experimental. We can already see the many creative possibilities of AI to problematically disrupt the workplace and augment capitalism's drive to lower costs via displacement of people. There are other ways that AI is being used to resist this and develop interesting alternatives. Some of these are explored in really interesting ways in Ruthanna Emrys A Half-Built Garden.
SIX: Anyone who pronounces definitely on such topics or the future in general is in for a shock. Over the next couple of weeks, we will experiment with some interesting directions in the emerging world of AI, knowing that the future is emergent and exceeds prediction. We are not that interested in making definitive claims about what it can do or how humans are inherently different (and better). Rather, we are curious about what is being done with it and how we can invent ways that these systems of Algorithmic Mimicry can augment various creative processes.
As we mentioned last week, here is something worth adding to your calendars – the 2024 Varela International Symposium: Sentience and Intelligence: AI, the More-Than-Human, and Us is happening in late May. As one of the key yearly symposiums on enactive practices, it is always interesting. This year, it is of special interest to us as they are turning part of their attention to the question of AI. The enactive approach to innovation has much to say on this as it strongly critiques and refuses the computational approach to sentience and cognition—without ruling out the sentience of artificial beings, etc.
The dates are May 24-26, and it is both virtual and in person by donation. Take a look and we hope to see you there.
We want to go back to the quote from Johannes Jaeger and highlight one key sentence:
“Finally, algorithms exist in a small world, in which all possible problems are well-defined, whereas organisms live in a large world, where most problems are ill-defined (and some are probably not properly definable).”
Much of our living activity and sense-making is not something that can ever be explicit and symbolic; it is tacit and inherently unspecifiable (we write about this at greater length here). We live and thrive in a world that is mainly unspecifiable. We are making sense—actually doing the creative work of making in every act—we make a path in walking. And this is not in any way the same as “interpreting” and “understanding"—life is far more creative than the act of reading off data. The NY Art Book Fair and the books we mentioned by Chang Yuchen and Audra Wolowiec are ones that refuse the acts of "understanding," and as such are wonderful reminders of how we thrive in the work of making the world – a making that is always excessive and fringing into the unknowable qualitatively new…
Till next week, stay hybrid, extended, artificial, creative, and more-than-human! (We are going to take a short, very human nap and then set off to explore Friday.)
Till next volume 139,
Keep Your Difference Alive!
Jason and Iain
Emergent Futures Lab
+++
📈 P.S.: If this newsletter adds value to your work, would you take a moment to write some kind words?
❤️ P.P.S.: And / or share it with your network!
🏞 P.P.P.S.: This week's drawings in Hi-Resolution
📚 P.P.P.P.S.: Go deeper - Check out our book which is getting great feedback like this:
