These videos I feel were the first to really give a glimpse into the weird and surreal future we're heading into.
The whole Balenciaga shtick felt like a metaphor for the rising tide of machine-generated content— and I tried to weave this narrative through the quotes of this computer history version, "the balenciaga revolution is far more important..." yada yada, while ostensibly about the history of computing, it is moreso intended to let us ponder its future.
I mean, fair. My first impression of the harry potter one was that it was some high-budget ad, but it's just some guy making weird mashups and exploring what's possible with AI. The balenciaga one just stuck because it was a perfect collision of familiar (harry potter) and uncanny (luxury fashion).
It might be the cynic in me talking, but I doubt this wave of Balenciaga ads is just the outcome of a random guy overusing a set of parameters. It would be trivial to find any other theme to exploit, but this is too focused on a single brand and relentlessly pushing it's name. It's guerilla marketing no matter of where you look at it.
Do you have any proof? Everything about it seems like it's a naturally-occurring meme. Nobody pushed it until it went viral 2 weeks after the original video, and the channel has clear history of many different experiments that didn't go viral.
This seems like "organic" marketing in the sense that an individual is parodying a brand but as a consequence giving the brand lots of exposure. This is no more guerrilla marketing than if a food reviewer gave exposure to McDonalds.
I wonder if you could make an AI that could generate likely meme ideas and then remove people from the equation completely. Seems pretty doable, if not already possible.
The problem is that memes, at least the ones that really take off, work because they are creative and funny, and so far that is probably the area AIs like GPT4 struggle the most with. It makes sense when you think about it, a technology that is built around the idea of predicting likely continuations will struggle with learning what is 'likely' in a context that asks for an element of surprise.
I.e., you'd probably need a pretty deep model if you want it to be able to understand that 'okay usually I combine likely concepts, but _now_ I need to combine unlikely concepts that still make sense together on a lower level'.
https://www.youtube.com/watch?v=iE39q-IKOzA
https://www.youtube.com/watch?v=ipuqLy87-3A
> Master has presented Dobby with Balenciaga. Dobby is... free.