The AI Ouroboros
The Ouroboros is considered a sign of immortal, longevity, the never-ending (story1) cycle of birth, destruction, and renewal. But the sight of it has recently been on my mind as I read, utilize, and come to terms with generative AI in our cultural, internet zeitgeist.
To get something out of the way: I don’t think AI is an artificial intelligence. It is a massive-in-scope, iterative machine-learning model that can approximate human thought and writing, but these approximations are copying and mixing known quantities.
Now, you might argue that much of creative thought is similar to this iterative approach. As humans, we remix much of what has come before - to new packaging, new ideas, jumbled and mixed together to create new unique ideas, art, writing, etc.
However, and especially from what I have seen in generative ‘art’, that final leap of creation toward a new unique approach hasn’t happened. Bits and pieces are formulated in hodgepodges of work, reminiscent of amateur stitching where seams still show through in any completed work. Hands with six fingers.2 Skewed perspectives. ‘Gooey’ features. Smeared faces. Nightmarish feats of illogical grotesque apparitions. The machine is desperately trying to manifest its own ghost. This leads to AI proponents or marketing teams to pitch their machine-learning services that don’t ‘hallucinate’. I never had to worry that MS Word’s Clippy mascot was going to hallucinate copyedits!3
A copy of a copy of a copy of a …
And so I come back to the Ouroboros, the infinite, twisting monstrosity devouring itself, and I wonder where this massive deluge of generated content will take us. It seems to me, that - as we rely more upon this technology - the machine is learning from itself across the internet: force-fed its own tail, firmly snared within its jaws and devouring the very content it once generated. How much human-made content will we see on the internet in 5 years?
Lost in Translation
…in what the researchers believe to be a ploy to garner clickbait-driven ad revenue, AI is being used to first generate poor-quality English-language content at a remarkable scale, and then AI-powered machine translation (MT) tools transcribe said content into several other languages. The translated material gets worse each time — and as a result, entire regions of the web are filling to the brim with degrading AI-scrambled copies of copies.4
Eat a Rock
Google’s implementation of AI in search has seen a disastrous rollout as well, with reddit sarcasm passing for thought leadership. From adding non toxic glue in pizza cheese (for cohesiveness of course) to eating a rock a day (it keeps the doctor away?), the search engine shows a clear misunderstanding of human humor5. Clearly, when AI is in charge of protecting itself from and filtering misinformation, it has shown a profound misunderstanding of reddit, which - to be fair - who hasn’t?6
At what point will this lack of discernment pivot from funny tweet7 responses, to a misinterpretation unleashing a nuclear holocaust on humanity over a misplaced comma? As we continue to hand over more and more ‘thought’ to machines, this will be the concern festering in my gut. And since I have to be anxious, I guess I figured I’d pass that along to you as well. 🙃
Additional Reading:
“When AI Is Trained on AI-Generated Data, Strange Things Start to Happen,” MAGGIE HARRISON DUPRÉ, FUTURISM.
https://futurism.com/ai-trained-ai-generated-data-interview
If you have seven fingers on a hand, I’m sorry… this is just to make a point.
There was that meme about helping people write suicide notes, which was a nightmare in its own right.
“Huge Portion of Internet is AI-generated Slime…” FUTURISM | LINK
human humor human humor human humor human humor… sounds funny when I say it - sorry.
It scares me on there. You’ve been warned.
YES, I CALLED THEM TWEETS. I WILL NEVER CALL THEM … THINGS? POSTS? TRULY AWFUL BRAND MALFEASANCE. BLURGH. SOME SUCH NONSENSE ABOUT HUBRIS AND THE FOLLY OF MAN. /END CAPS