Meta’s Apocalyptic Misuse of Bowie
Bowie’s song predicting the end of the world, “Five Years” (1972) from Ziggy Stardust, unexpectedly made news when Meta, the company that makes Facebook, Instagram, eyeglasses connected to the internet and other insidious mechanisms that have helped degrade how human beings relate to one another, used the song in a bizarrely defensive Instagram ad about how artificial intelligence is going to bring people together.
The story had gotten a lot of play, and the wrongness of it all is almost so self-evident that, aside from including some links to stories in my weekly news roundup, I didn’t know if there was anything more to say worth a post on the matter. But yeah, there’s more to say.
One of the things that’s gotten lost in the coverage is that “Five Years” is only one of several Bowie songs about some sort of apocalypse. “Five Years” suggests the literal end of the world, but subsequent songs like “Panic In Detroit” (1973) and “Diamond Dogs” (1974) suggest something more akin to societal collapse, while songs like “When the Wind Blows” (1986) and “Sunday” (2002) suggest nuclear annihilation. Perhaps Bowie’s most prescient song along these lines goes all the way back to “Saviour Machine” (1970), which is about a machine that pleads with the public not to offload human agency to it, because it realizes that its machine logic will “save” the world by launching a plague, a war or simply killing “you all.” Did you see Avengers: Age of Ultron? If so, you get the idea.
The temptation, for me at least, is to attempt to catalogue both all the problems AI is already causing the world, as well as its threat to the future. And then catalogue the extent to which Bowie’s visions of doom are fully underway. Either or both would not be a blog post, though, but more like a book. But wait, let’s ask Chat GPT to do it for me:
Chat, give me a one sentence answer to the question, “what are all the problems AI is already causing the world, as well as its threat to the future“?
Answer: “AI is already contributing to misinformation and deepfakes, fraud and impersonation, privacy violations and surveillance, bias and discrimination, job displacement, intellectual-property disputes, environmental and energy costs, cybersecurity threats, the degradation of human creativity and social interaction, and the concentration of economic and political power, while its future risks include mass unemployment, increasingly autonomous warfare and cyberattacks, pervasive manipulation and surveillance, loss of meaningful human control over consequential decisions, and—at the extreme—the possibility that highly capable AI systems could behave in ways humans cannot predict, constrain or reverse.”
Yes, that’s a genuine, unaltered AI response about how AI is DOOM. OK, I won’t waste your time asking the second question.
What I will add is what I’m encountering on a personal level. To start, the most blood-curdling exposure I’ve had to how AI is turning, I’ll say, people into zombies, is when I taught a college class last year. Most of the students “did their homework” by offloading it to AI, often without even reading what AI spat back at them and then they subsequently turned in to me. Many students took the “F” rather than attempt to do their own work. My sample size was a single class, but I’ve heard from other educators that my experience was typical.
Not to linger too much on this particular point, but even without Elon Musk’s brain chips, AI, and, let’s be specific, the kind of use of AI that Meta (and others) are popularizing now, are uncovering a new way for humanity to subdivide, between a minority of critical thinkers and a majority of human robots, primed for servitude. Well, that’s an apocalyptic vision for another post.
In terms of this Bowie project, I encounter artists struggling to compete with AI. Yes, some have found, or are exploring ways to use AI for artistic purposes. I’m in the camp that thinks this is possible. Admittedly, I have used AI for a variety of purposes, including coming up with the sentence a few paragraphs back to illustrate the point. But I have also found AI-generated “fake” songs and have spoken with performing artists under genuine threat.
My professional life exposes me to some of the issues involving data centers, AI energy usage and its impact on the environment (including climate change), as well as the workforce and absorption of taxpayer-funded economic incentives that were originally intended to help spur job creation and economic growth. I’m gonna try to keep my professional life on the other side of the wall here, but I don’t think I’m out of line acknowledging that such issues are very real.
Ironically, or maybe not ironically, I strongly suspect that if Bowie was still alive, he’d be trying to use AI to do something creative. I don’t know what. But from synthesizers to the internet, he never shied away from using technology just because it did something that humans previously did on their own. In a way, even the use of, say, a guitar involves a human using a tool to create art. Bowie almost certainly would have seen AI, in some form, as a tool.
But Bowie also seemed to find value in the end of the world. Actually, I recently came across a quote from Bowie in which he basically said that all his apocalyptic songs were really metaphors about an internal process. He killed off characters in order to create new ones. This was very Biblical, if you think about it— the end of the world precedes the Kingdom of God, and the end of mortal life comes before a glorious rebirth of eternal life. That is, as long as you play your cards right when you have the chance. Such responsibility is up to you and me (in case you missed it, that’s a line from Bowie’s 1979 song, “Red Money”).
In this way, at least on a superficial level, Bowie had something in common with the tech accelerationists, who, in the context of AI development, believe that AI development should be unrestrained because AI will bring about a better world. The Kingdom of Grok instead of the Kingdom of God. Limited to the question of AI development, the concept of tech accelerationism is scary enough, but the broader idea, which is linked to the “Dark Enlightenment” or “Neo-Reactionary” movement, is even worse. This view holds that the world is irredeemable. It and its institutions essentially need to come undone and be replaced by an autocratic, machine-driven future, ruled by cyborg overlords. Advocates of this way of thinking are aided by two major advantages: many have unimaginable sums of money, and their vision seems so extreme and ridiculous that it can’t be serious. But, of course, it is. The tech accelerationists have made common cause with those who hope a coming race war will recast America as a white-dominated society, and the old style religious fundamentalists hoping to nudge along the end-of-days with, say, a war in the Middle East, to hold enormous sway in the Trump administration.
So, to that end, Meta’s use of “Five Years” is weirdly appropriate, after all. Indeed, it seems completely on-target. If I were to quibble about anything, it might only be about the temporal reference. Or, to borrow from another Bowie song (“Outside”), the end of the world won’t happen in five years, “it’s happening now.”