What YouTube Hustlers Can Teach Us About AI

Photo-Illustration: Intelligencer; Photo: @SmartMoneyTactics/YouTube

The tech industry is all-in on AI. Tech giants are pumping massive resources into research and new products. Microsoft and Google are suggesting they’ll revamp their entire product lines. AI start-ups have collectively raised tens of billions of dollars in the past year, during a tech slowdown. The buzz has broken Silicon Valley containment, as regular people have had the chance to interact with new and surprising tools. Over a hundred million people have played around with ChatGPT, and millions more have tested out media generators such as DALL-E and Midjourney. OpenAI’s GPT-4 model, released this month, is turning napkin sketches into functional websites and explaining memes. Lots of people are coming away dazzled. Others are remaining skeptical.

Practically everyone’s got a hunch about where this is headed next, but nobody knows for sure. Is AI coming for our jobs? Will it remake the economy? Is this yet another in a string of spectacular but ultimately disappointing hype cycles or the beginning of something much bigger and weirder?

There is one source, however, from which enthusiasm about new AI tools is absolute and unconditional — less conflicted, even, than at the firms actually creating them: the online hustlers. Across the loose economy of popular influencers who specialize in passive-income schemes, hustle porn, work-for-yourself tricks, and motivational content, the embrace of new AI tools has been rapid, enthusiastic, and complete. Earlier this year, James Vincent of The Verge noted ChatGPT’s momentous arrival on the scene. “On YouTube, Instagram, and TikTok, a motley assortment of established and would-be financial influencers are pumping out videos advising how you — yes, YOU — could be making tens of thousands of dollars in your sleep with the help of OpenAI’s chatbot,” he wrote.

Photo: John Herman

It’s become a bit of a meme. Last week, designer Jackson Greathouse Fall set out to see just how much of the hustle he could outsource to GPT-4 by asking it to create a new company for him — a challenge he called HustleGPT:

On these dilapidated outskirts of the internet economy, where the pursuit of money is uncomplicated, content is just content, ethics are for suckers, the AI revolution can’t happen soon enough. “We can consume ten times the information and produce ten times the output,” says YouTuber Alex Hormozi in a video titled “Get RICH in the A.I. Revolution (2023),” “which is a 100x increase.” His vision for AI — it “has the potential to unravel the very fabric of capitalism itself,” he says, and could “eliminate the workforce entirely” — is dystopian for everyone except his million followers, who will be set up to thrive. People who put their heads in the sand, he says, are doomed. But savvy opportunists who embrace AI tools, as he plans to, will flourish. It’s a world in which a budding entrepreneur with a side gig can suddenly “do a day’s worth of work in a few minutes,” making more money and freeing up time. In the coming AI apocalypse, the risers and grinders won’t just survive — they’ll finally be raptured into passive-income heaven.

It’s not much of an overstatement to say that hustle culture has adopted AI in almost religious terms; it is, in the broadest possible sense, a faith-based community led by charismatic preachers. Most online hustle influencing, as described by Rebecca Jennings at Vox, draws from a long tradition of self-help gurus who make money by “telling other people how to replicate their own business successes” while cultivating their inner selves. Where the previous generation sold books to sell seminars, the neo-gurus build profitable YouTube channels to sell … seminars, but online.

Their theology, such as it is, combines familiar ideas about working harder and smarter with an emphasis on a very particular sort of masculine self-improvement and a generally ruthless view of other people, who are either predestined to wage slavery or deservedly stuck there as a result of their own laziness. (It takes a dim view of their work as well, which is underoptimized and not sufficiently focused on the only acceptable reason for doing it in the first place, which is making money.) Its exhortations to try harder and be better than everyone else are interspersed with, basically, hacks: shortcuts and secrets with which to beat and ultimately escape the system.

Photo: John Herman

Services like ChatGPT also give lower-end hustle influencers a lot to work with. Tools for generating commodity content sound like a perfect match for some of the most common income schemes in hustle culture, about which you can watch hundreds of instructional videos: drop-shipping operations marketed with floods of SEO and social-media content; harvesting commissions from affiliate programs by, again, flooding the internet and social platforms with content; becoming a hustle influencer yourself by, of course, posting lots of content. Other AI tools — for image generation, photo editing, translation, and programming — promise to turn every side-gigger into a hyperproductive generalist freelancer, able to whiz through programming jobs on Upwork, knock out art commissions on Fiverr, or do translation work on sites like Guru, raking in “$100 a day” or “$6,147 a week” or “$300 PER VIDEO.”

That most of these schemes don’t work most of the time didn’t matter before, and it doesn’t matter now that they use AI. ChatGPT, a tireless companion that promises to do the grinding for you, makes the story sound more plausible and shores up the fantasy.

When Greathouse Fall asked GPT-4 for a business plan for the company he was letting it create, it extracted from its training data precisely the sort of paper-thin scheme that has sustained low-end hustle content for years:

The experience of playing with AI tools such as ChatGPT and DALL-E — which really can produce imitations of email-marketing copy and mass-generate Instagram posts and write scripts that sound as if they were written by popular YouTubers — provides the temporary sensation of being in charge not exactly of someone but of a someone-like thing. In a community of temporarily embarrassed millionaires and bosses-in-waiting gathered together to consume content that gives them a sense of control over their futures, a text box that does as it’s told counts for a lot.

Photo: John Herman

The sudden arrival of AI as a culturewide fascination — and as a technology that more sober people are both hyped and worried about — has also led some more ambitious hustle influencers to slip into a millenarian frame. “People predict AI will, in the not-too-distant future, have a level of intelligence that isn’t just smarter than the smartest human — it will be smarter than all humans combined,” says Tom Bilyeu in a video called “Use AI to Get Ahead While Others Panic (PREPARE NOW).” Bilyeu, a motivational influencer (the channel’s previous video is titled “The TOP FOODS You Must Eat to Lose Weight & END INFLAMMATION”) who made his fortune selling a protein-supplement company called Quest Nutrition, invokes Ray Kurzweil’s concept of the technological singularity (“The moment when technology advances so fast that we can no longer predict where things are headed”) and warns that AI can already “improve itself.” It’s all pretty hopeless. AI is “laser-focused at all times,” he says, and its cognitive abilities are “faster than yours or mine.” The only way to survive the future, he warns his millions of followers, is to be like Elon Musk and “invent it.”

There are limits to what we can learn from hustle culture’s early embrace of AI — we’re talking a loose group of people who don’t represent an industry or a type of work so much as a subgenre of influencer. They’re storytellers sketching a caricature of new technology from within a parody of the economy. The plans they suggest to their followers are the sorts of things that are first in line for full automation. They’re in it for the followers, and their followers are in it for the story — otherwise, they would have stopped watching long ago.

It took less than a full week for Greathouse Fall’s HustleGPT experiment to isolate some related truths about the hustle. Its plans were dull, and its advice was bizarre — recently, it suggested hiring a couple of human copywriters — but within a few viral days, that was well beside the point. It was a great story. At last count, HustleGPT had raised thousands of dollars from “investors” on Twitter who were charmed by its story, reached millions of people on social media and through news coverage, and recruited more than a thousand people into a paid Discord channel for those who wanted to try out the challenge for themselves.

The hustlers do have something to tell us. As they figure out how to message AI to their audiences of frustrated young men and bored office workers, settling on a combination of fearmongering balanced with wild-eyed individual optimism, voices at the highest levels of AI research are themselves speculating wildly about what their tools may mean for the economy, oscillating between optimism and fear, experimenting with uncertainty and confidence, for an audience that includes curious Twitter followers and newly anxious C-suiters alike. Each has a compelling and valuable story. AI’s vague, just-over-the-horizon promise of infinite productivity suggests different and perhaps conflicting things to different people: to the hustler, a reduction of his own labor; to the CEO, it’s the reduction of paid labor. These are dreams, and Sam Altman, head of OpenAI, speaks to both:

If you spend enough time watching hustle-culture videos, you notice some patterns: Big up-front promises are rarely kept; specifically, as they are disclosed, plausible-sounding plans are revealed to be incomplete with crucial missing steps and magical assumptions. The YouTuber, who is already successful (or who is pretending to be successful), clearly hasn’t tested them out and inevitably hands back responsibility to the viewer — these plans will work if you can execute. At best, they only know how they got rich; at worst, they’re misleading you for their own gain.

Likewise, the tech companies building these new AI products, which are fresh, well funded, and barely monetized, can only guess, not unlike an LLM, what is statistically most likely to come next in this sequence for the rest of us. It’s easy to confuse them with keepers of secret knowledge, but they don’t know where all this is going. For now, they’re telling stories, and lots of people are listening.