Good Lord, where do we start?
The narrative driving AI these past few weeks has been operatic, and true to form, it’s building toward a frenzied crescendo. But what kind of opera are we living inside? Is it Tragedy? Semiseria? Farce?
Perhaps it’s all three.
To review, we’ve recently learned that “swarms of rogue agents” have been attacking various government and commercial sites for months. Despite that somewhat disconcerting news, both OpenAI and Meta introduced cuddly “personal agents” that they promise – really, really, they promise! – won’t ingest your personal data and feed it to their surveillance-driven advertising machines. Our Fearless Leader got into the mix by proposing a new shade of lipstick for the category’s brand, which thanks to the nearly $1 trillion in “new wealth” that AI billionaires have garnered, certainly feels a bit porcine*. Meanwhile, an escalating Greek chorus of skeptics are screaming that the world economy is teetering on the brink of an immense AI-driven crash.
And those are just the biggest headlines of the last month. Never mind that the AI race is now framed in the narrative of a “religious war,” or that a leading oracle of the industry tells us we should be cool with “some bad things” happening as a result of all this madness. Airplanes crash sometimes, he points out, but hey, we still get on airplanes! (Ok, it’s true we don’t have the FAA making sure those planes don’t drop out of the sky, but hey, that’s OK, everyone pinky promises they’ll keep things under control.)
As we attempt to digest all of this, we’d be wise to question the core values of the industry that has gifted us this narrative.
For more than two decades, the tech industry’s motto has been “ask for forgiveness, not for permission.” This philosophy has allowed entire generations of startups to grow “in the gray,” executing business models that leverage legally murky waters around copyright, personal data, monopolies, spamming, and more. I first noticed the practice during the rise of “Web 2” – spammy growth-hacking tactics built briefly ascendant winners like Zynga and Demand Media, while more ambitious upstarts like Uber and AirBnb brazenly flouted regulations meant to temper the worst instincts of industries like transportation and hospitality. Regulations? Those are for the old economy! Here in the new, we have no need for your silly rules!
All of this was driven by the mantra of “blitzscaling” – getting as big as possible, as fast as possible. If you achieved a critical mass of customers before the money ran out, you could bulldoze limpid politicians into submission, winning riches and fame along the way. It’s not a coincidence that Meta’s official motto is “move fast,” a shortened version of the company’s original credo: “Move fast and break things.”
This “go big or go home”mentality pervades the Valley ecosystem – VCs push it, Stanford teaches it, the press canonizes its success. And nowhere is it more ingrained than at large platforms like Facebook/Meta, where “growth hacking” was invented, or OpenAI, which is run by a man whose previous job was minting growth-hacking companies through the venture incubator Y Combinator.
Classical operas follow a predictable narrative arc. First is the exposition, where we meet the characters and the context in which they operate. Then we have a period of rising action, where the story begins to build. Next is the climax – a turning point where consequences become clear, and our protagonists’ fate is sealed. This leads to the hero’s aria, where we collectively contemplate the implications of a hero’s fall. And finally is resolution, often called the denouement – if the opera is a tragedy, this final act is usually referred to as the Catastrophe.
Keep that in mind when you read about those “rogue agents” OpenAI and others have released into the world. And notice the language – the word “rogue” implies an independent character acting outside institutional control. “Rogue” externalizes the negative consequences of technology, making them society’s problem, not the Valley’s.
But these AI agents are in no way rogue – they’re doing exactly what they’ve been programmed to do. As the music swells and our heroes begin to sing their arias, we’d do well to remember that their actions are the result of explicit business practices that Valley companies have perfected over the past 25 years.