Earlier this week Belinda Ridgewood posted some fascinating examples of AI-generated video, with images culled by artificial intelligence to represent actual song lyrics (I missed the post until today). On the same subject, Benj Edwards from ArsTechnica, reports that, thanks to new open-source AI image modeling, in only a year or two the internet will be saturated with AI-generated images, including purportedly “genuine” images of people indistinguishable from the real thing.
AI image generation is here in a big way. A newly released open source image synthesis model called Stable Diffusion allows anyone with a PC and a decent GPU to conjure up almost any visual reality they can imagine. It can imitate virtually any visual style, and if you feed it a descriptive phrase, the results appear on your screen like magic.
I sometimes wonder what folks from ancient history actually looked like. Socrates, for example was said to be rather unattractive (according to one source “He had wide-set, bulging eyes that darted sideways and enabled him, like a crab, to see not only what was straight ahead, but what was beside him as well; a flat, upturned nose with flaring nostrils; and large fleshy lips like an ass.”). But it now appears that thanks to AI our days of wondering about such things are numbered. For example, here is an actual photograph of the ancient Greek philosopher, Heraclitus, amazingly taken in 500 B.C. And here are some beautiful landscape photos suitable for Facebook sharing (when your own mundane photos aren’t generating enough “likes,” I guess). As Belinda’s post noted with its Van Gogh-esque AI rendering of Kitchen Table Kibitzing’s header photo, art, too can be generated to look exactly as if a human being created it.
Like most technologically-based developments these days, open source AI modeling implicates some serious societal, artistic and ethical issues. Edwards shares this tweet:
Stable Diffusion's public release has raised alarm bells among people who fear its cultural and economic impact. Unlike DALL-E 2, Stable Diffusion's training data (the "weights") are available for anyone to use without any hard restrictions. The official Stable Diffusion release (and DreamStudio) includes automatic "NSFW" filters (nudity) and an invisible tracking watermark embedded in the images, but these restrictions can easily be circumvented in the open source code. This means Stable Diffusion can be used to create images that OpenAI currently blocks with DALL-E 2: propaganda, violent imagery, pornography, images that potentially violate corporate copyright, celebrity deepfakes, and more. In fact, there are already some private Discord servers dedicated to pornographic output from the model.
{***}
Realistic image synthesis models are potentially dangerous for reasons already mentioned, such as the creation of propaganda or misinformation, tampering with history, accelerating political division, enabling character attacks and impersonation, and destroying the legal value of photo or video evidence. In the AI-powered future, how will we know if any remotely produced piece of media came from an actual camera, or if we are actually communicating with a real human?
The tension between AI modeling “art” and genuine artistic talent reminded me of something I’m really looking forward to this week: the premiere of Brett Morgen’s film, Moonage Daydream, about David Bowie.
As reported by Melena Ryzik for the New York Times, the film opens Sept 16th to the broader public.
Morgen’s opus about Bowie, “Moonage Daydream,” which opens in theaters and IMAX on Sept. 16, is billed not as a traditional documentary but as an immersive experience. It’s equal parts psychedelic and philosophical — a corkscrew into Bowie’s carefully constructed personae, assembled entirely from archival footage and audio, some of it rare and never broadcast before. The effect is “a hallucinatory jukebox doc with killer subtext,” as one reviewer wrote, appreciatively, after it premiered at the Cannes Film Festival this spring.
[***]
Tony Visconti, Bowie’s longtime collaborator and producer, who served as a resource for the audio, came away impressed with the way the film kaleidoscoped the visuals, narration and music. “There is technical wizardry in all that,” he wrote in an email. “And when seen and heard, especially in an IMAX theater, you will get the most Bowie ever — sensory overload.”
I’m taking my son to see it on IMAX. The trailer itself is pretty impressive:
After spending five years making a film about such a cultural icon, Morgen came away impressed most with Bowie’s sheer humanity:
For Morgen, one of the most illustrative points was the way Bowie behaved in many interviews, often with people who clearly did not get him; one, trying to suss out just how alien this gender-bending artist was, asked if he’d had a teddy bear as a child. And yet, “I never saw David talk down, be disrespectful, short, annoyed,” Morgen said.
Maybe this was just politesse as a disarming tactic, but Morgen saw it as something deeper — an ability to seek connection and profundity in any situation. It was a message that he tried to convey in the film. Bowie was “trying to make each moment matter,” he said. “It’s a life-affirming sort of road map, on how to lead a satisfying and complete life.”
Bowie weighed in on the impact of the internet (and implicitly, AI) back in 1999, calling it “an alien life form:”
As transcribed by The Teeming Brain:
I don’t think we’ve even seen the tip of the iceberg. I think the potential of what the Internet is going to do to society, both good and bad, is unimaginable. I think we’re actually on the cusp of something exhilarating and terrifying.
Paxman: It’s just a tool though, isn’t it?
Bowie. No, it’s not. No. It’s an alien life form. [Laughs] Is there life on Mars? Yes, it’s just landed here.
Paxman: But it’s simply a different delivery system. You’re arguing about something more profound.
Bowie: Oh, yeah. I’m talking about the actual context and the state of content is going to be so different from anything we can envisage at the moment, where the interplay between the user and the provider will be so in sympatico, it’s going to crush our ideas of what mediums are all about.
This was a man so far ahead of his time, we’re still a long ways from catching up to him.