Critical Loop Literacy

We need something like 'loop literacy.'

You could call it systems thinking, but these days that seems to descend too quickly into contextless canvases for consultants.

Being able to see self-reinforcing loops is especially important for the criticality needed to stop us being duped by AI Snake Oilers (hat tip to https://press.princeton.edu/books/hardcover/9780691249131/ai-snake-oil#preview).

Expectation Management

For example, humans are smart because they generously project other minds into what they interact with, whether implicature maxims required for conversation with other people (https://www.wikiwand.com/en/articles/Implicature), but also any attempt to negotiate anything living which is presumed to have some kind of consistent intent (animism). We presume things are intelligent, and then make them seem intelligent by intelligently cooperating with things to get to intelligent outcomes. To pervert the title of Frans de Waal's book (https://wwnorton.com/books/Are-We-Smart-Enough-to-Know-How-Smart-Animals-Are/), we are too smart to know that AI is not smart, only made smart by our projecting collaborations with it.

(See relatedly, Shannon Valor's The AI Mirror: https://global.oup.com/academic/product/the-ai-mirror-9780197759066.)

Gen AI UX = Gen AI of a Gen AI + Gen AI of another Gen AI

Actually, it is less about us being smart or not and more about Generative so-called AI interface design being deliberately deceptive about how many loops those systems are deploying. Uncritical users think they are interacting with one super-smart system when there is only the mimicry of intelligence as a result of a seemingly unfathomable set of interactions between a series of systems. For example, consider this from an ArsTechnica article (https://arstechnica.com/information-technology/2024/09/openai-threatens-bans-for-probing-new-ai-models-reasoning-process/) on attempts to 'prompt inject' reveal those several loops:

"When users ask an "o1" model a question in ChatGPT, users have the option of seeing this chain-of-thought process written out in the ChatGPT interface. However, by design, OpenAI hides the raw chain of thought from users, instead presenting a filtered interpretation created by a second AI model." [my emphasis]

You can have any colour you want from this range of one colour

These loops are not just social or technical. They are of course primarily the self-reinforcing logic of capitalist markets. Zuckerberg, whose brand equity has recently been unjustifiably inflated by Musk's pathological immaturity, has just pronounced:

“Look, we’re a big company,” he said. “We pay for content when it’s valuable to people. We’re just not going to pay for content when it’s not valuable to people. I think that you’ll probably see a similar dynamic with AI.” https://www.theverge.com/2024/9/25/24254042/mark-zuckerberg-creators-value-ai-meta

It is of course hard to value what you did not know existed, and it is only in the tightly looping small minds of profiteers that absence of evidence is evidence of absence.

(See relatedly Jacob Ward's The Loop: How Technology is Creating a World without Choices: https://jacobward.com/theloop/.)

I am constantly berating students for example for how impoverished their filmographies are. According to Gen X me, it impossible to be a designer without a very diverse visual palette such as obsessive cinematheque viewing affords. However, on the odd occasion that a student takes up my suggestion to watch a very cancel-able Peter Greenaway film, they invariably get in contact with me to say that it is impossible to find, even on old school torrent sites.

The least job of higher education is to expand students' loops

The wealthy who have over-invested in generative so-called AI platforms are only going to make their money back and reinforce their ecosystem depleting power by ensuring that we all only like the slop these systems serve up. They are playing the same game that neoliberalism did with government services: under-resource them so people cannot find value in them and then announce that those people 'evidently' want privatised, digitally transformed, DIY options.

In the face of this conspiratorial loop, it is essential that whoever can, expose people to as diverse a set of 'content' as is possible. Universities for example, would seem to have a duty to, now more than ever, show students authentic creations from across histories and cultures in order to expand their sense of what they might like - and, more importantly, what they might value even though might not like it.