🏢 Beyond corporate-first AI imaginaries

Throughout 2023, we in the AIxDESIGN community found ourselves increasingly disillusioned with the prevailing discourse and narratives around AI.

The central narratives around emerging AI technologies today are being shaped by Big Tech companies such as Microsoft, Google, and Meta – with for-profit products such as ChatGPT dominating our idea of what this technology is, can be, and most likely will be in the future. In some ways, “Silicon Valley’s most powerful monopoly may be how we perceive technology.”

GROWTH

These dominant AI narratives and development are heavily influenced by the interests of big tech companies - primarily growth, profit, and stakeholder value, the usual capitalist objectives. It is sometimes said that “the problem with AI is the problem with capitalism”.

IMG: Ben Grosser’s redaction poetry version of Andreessen’s Techno-Optimist Manifesto

TESCREAL

Beyond that, there’s the underlying ideologies and beliefs of people leading and funding said developments. Their taglines of “AI to benefit all of humanity” are intended to sway the public in its favor with ideas of ‘a promised land’ as this VOX article sharply notes sounds a lot like ‘religion, repackaged’. A recent paper by Timnit Gebru and Émile P. Torres bundles some of the underlying ideologies driving the agenda towards AGI (artificial general intelligence) in particular into the TESCREAL acronym, clearly labelling these beliefs and tracing their origins back.

Silicon Valley’s vision for AI? It’s religion, repackaged.

The TESCREAL Bundle

DISCOURSE

While there’s promises of a better tomorrow, there is at the same time a lot of fear-mongering and panic circulating of AI ‘replacing us’ or ‘going rogue’ as well as a growing awareness of the actual harms and risks of AI displacement, bias, and environmental impact to name a few.

Struggling to reconcile these two stories and the nuance in between, the discourse around AI, both in media headlines and day-to-day conversation, often falls into polarizing positions:

👍 with techno-optimist / hype / utopia on the one hand

👎🏻 and doomers / criti-hype / dystopia on the other

This binary framing of two warring visions isn’t fertile soil for constructive conversations, leading people into false debates, and fails to capture the nuanced and complex realities of how AI is shaping our lives in both utopic and dystopic ways all at once.

Not resonating with either of these options, we try to practice a third position that exists in the in between; acknowledging a plurality of (at times, juxtaposing) truths, and focusing on our agency within them.

IMAGINARIES

This media coverage, propaganda, and discourse shape people’s perception - affecting both experts and laymen alike.

Serpentine’s Future Art Ecosystem annual briefing on Public AI also outlines that “the narrative wars around AI aren’t just about words and concepts. The ways in which AI is framed and which narratives will stick in the popular imagination are deeply entangled with the ways in which the technical elements of AI will interweave with broader societal dynamics”.

The more academic term ‘sociotechnical imaginaries’ equally is about into “the societal understandings of the possibilities, potentials and risks embedded in the technology” and **how it “plays a large role in influencing how AI is perceived and engaged in societies”.

The narratives we tell about AI matter because the language and stories we share > shape our beliefs > shape our imaginations > then our actions > then our realities. Reverse engineering, this also means that telling different stories can bring forth new realities because “everything that is real has to be imagined first”.

Dreamscapes of Modernity | Are.na

In simple terms, we acknowledge “words make worlds”. To quote the more eloquent Donna Haraway “It matters what matters we use to think other matters with; it matters what stories we tell to tell other stories with; it matters what knots knot knots, what thoughts think thoughts, what descriptions describe descriptions, what ties tie ties. It matters what stories make worlds, what worlds make stories”.

Instead of working within the symptoms and bounds of bigger influences, we felt a strong need to address the ways we think and talk about AI, and summon alternative imaginaries which can in turn bring forth new approaches and realities.

🌱 Seeding alternative narratives

The Slow AI project seeks to subvert this corporate-first thinking that often dominates AI narratives and approaches, and instead explore how values of mindfulness, care, and community can be applied to the development and use of these powerful technologies.

A guiding principle for us at AIxDESIGN is this quote by Ruha Benjamin that reminds us to “imagine and craft the worlds you cannot live without, just as you dismantle the ones you cannot live within".

If the visions big tech is painting for us don’t resonate, what are the other possibilities? If that isn’t the world we wish to live in, what is? What are the alternatives to corporate-first AI?

Moving beyond critique, this work focusses on finding (learning from those doing this work already and those who came before us), imagining, and putting forth alternatives because we agree with the Solarpunk line of thinking that “the future of humanity needs a specific Imaginary to build towards as opposed to simply preventing catastrophe”.

We are committed to the active imagination, practice, exploration, and construction of these alternative paths for AI.

By creating counter-perspectives that are less binary, more nuanced, and open to “weird” and novel ways of seeing, we hope to spark a different set of ideas, and in turn; realities around AI.

We don’t aim for a prescriptive solution, but rather a kaleidoscopic plurality of views.

We’re seeing three non-linear parts of this process. It begins with unpacking the existing narratives; an exercise in disentangling the technology from the capitalist conditions in which we most often see it applied. What follows is an expansion of our perspectives through collective imagination, learning from those who came before us and people already practicing otherwise, and seeding alternative narratives. And finally, exploring how those ideas can bring about new approaches and realities.

The research encompasses both the exploration of emerging alternatives in thinking and the practical application of those ideas. The questions we’re asking:

🐌 The Slow AI framework

The term “Slow AI” emerged from grumblings within the AI research community, though its exact origins are unclear. We first stumbled upon it in this article on Timnit starting DAIR, but it’s unclear whether it’s something she said or the editor Eliza Strickland came up with.

Wherever it came from, it stuck with us ever since and found it a powerful shorthand for people to quickly and intuitively grasp what the project is about.

Latching onto the language around counter-movements like slow fashion and slow food, it positions the work outside / beyond the dominant and familiar context of ‘Fast AI’.

Just as slow fashion and slow food prioritize quality, locality, and sustainability over mass-produced, generic, over-engineered products, Slow AI seeks to apply these values to the development and use of AI technologies.

We’re not attached to words as we are to concepts, and have adopted Slow AI a resonant framework for seeding, collective referencing, offering our own interpretations, and inviting others to expand.

🪸 Embarking on the [participatory] research

Looking for funding to make this work happen, thankfully the Stimuleringsfonds grant shared our sense of importance (YAY) and supported us with (partial) funding that enabled the Slow AI project into existence.

To kick off the research, we hosted an initial session with the community to explore a range of potential directions in January, and through those conversations choose the three lenses of Small, Esoteric, and Ancestral AI.

Over the next months, using the AIxD playbook of participatory, community-led research, the Slow AI project will invite critical AI researchers, designers, creative technologists, and artists to explore / interrogate and publish critical AI discourse in formats that make sense to us and our practices such as involving participatory research sessions with the community, zine-making workshops with Internet Teapot, artist commissions for critical making projects, and open-source publishing.

We will pour our insights to create one hot compost pile of ponderings, miro boards, and zines, ultimately compiling everything into a final anthology to be released in fall 2024.

You can stay in the loop on where we are and ways to get involved through the Slow AI public project page, read about either of the streams - 🐜 Small AI, 🐚 Ancestral AI, or 🔮 Esoteric AI - or dig deeper into the methodology and approach here.

By collectively researching and practicing alternative ways of understanding & approaching AI, we hope to empower a global community of practitioners, thinkers & makers to join the critical AI discourse and lay claim on their visions presents and futures alongside AI.