Is OpenAI just Uber all over again?
Some thoughts on how OpenAI is just the Uber monopoly scam all over again
Uber was a technology that at the time it initially launched was pretty cool, at least technology wise. Taking advantage of the explosion of computers in our pockets and purses that knew where we were and could match us with drivers because it knew where they were was innovative. But the concept, drivers in cars, directed to locations where customers wanted a ride, had existed for at least 50 years prior, when taxi companies started using radios to dispatch cars to riders. So while there was a new layer of convenience, you didn’t have to be by a landline phone to get the taxi to show up in front of you and you could pay from your phone instead of carrying cash, it wasn’t actually a new idea.
What Uber actually represented was an attempt to monopolize ride-share, and to kill off laws that required licensing programs for taxi companies and drivers. In fact when Uber launched in many locations, it was illegal for them to be operating. They also skirted insurance regulations for many years, since the driver’s personal insurance didn’t cover them driving as a business, and Uber wasn’t interested in insuring the drivers because they were “independent contractors,” the grand colloquialism of late-stage capitalism. But the gig economy (ugh, can we go back in time and just not) was hot and investors remain eternally stupid so Uber got tons of money to do illegal stuff with an unclear business plan. Which was the key component of Uber’s success, though really it’s more like barely staying ahead of the jaws of failure than “success.”
Uber had a ton of cash, so they didn’t actually have to charge the rider the actual cost of a ride, which the free market of taxi companies had largely already established. Oh wait, no. Taxi companies, themselves bordering on monopoly and loving to collude with one another, had their rates carefully regulated by local government to ensure they weren’t over charging customers. Throw another law and/or norm on the bonfire of capitalist greed. But anyway, back to Uber, who undercut the price of taxis and other for hire vehicles to draw in customers. There’s an old joke my dad used to use regularly. “Lose money on every one, but make it up on volume.” It doesn’t work for a mom and pop shop or a corner store, but it definitely works for technology startups with stupid investors.
Amazon famously lost money for a decade while it drove local bookstores out of business. Uber, founded in 2009, and public since 2019 first turned a profit in 2023. Do you think you could find anyone to give you hundreds of millions of dollars to burn for 14, or even 4 years? Yeah, me neither. But Uber accomplished its goal. People open apps now, they don’t call taxi companies. And up came the prices. While the drivers almost certainly don’t keep enough of the ride costs to cover their gas, maintenance, and time, the company now has lock in and can skim enough off the top of every ride to turn a profit.
Which leads us to OpenAI. No one wanted to chat with websites. It’s actually an incredibly in-efficient way to get information. Sort of like the pivot to video where what could have been a Tweet now is a 10 minute video with 7 ad breaks and you only get to the answer in the last 30 seconds. Oh, but now half the information chatted back at you is a vast hallucination, good luck figuring out which half. And chatting with websites is incredibly expensive, both in dollars and in climate costs. So much so that OpenAI loses money on every customer, including the paying ones. But investors, still dumb, are shoveling money into the furnace, quarter after quarter. Because they can see the monopoly. If students can’t write papers without chatting, they’ll have nowhere else to turn. If professors can’t figure out if students cheated without ChatGPT, how will they grade? If our jobs have been taken away by the very bots we were befriending, how will we have a compelling resume or cover letter without the chat bots helping us.
But there is no there, there. It’s a technological party trick. It’s the app instead of the landline phone to call the taxi company. Students used to pay other students to cheat. Companies used to pay their employees to write stupid email content no one ever reads. Now we write emails using AI that the recipient summarizes using AI because they get so many damn emails because AI email represents no time investment to write it. AI companies have stolen all of the creative output of entire generations but scoff at the idea there could be a future without them. Investors know the only route to profitability is to never pay for all the content that their bots now mindlessly regurgitate. Two parts shitpost, three parts hot take, mix and … profit! But they invest anyway. We chat because the AI bros know from actual humans’ research that it makes us infuse their dead, soulless, planet destroying data centers with a humanity they do not, and will not ever, deserve.
Writers wrote because they loved it, artists painted because it brought them joy. Or at least tamped down some of their darkness. But AI doesn’t feel, it doesn’t learn, it just burns up the planet, one pixel at a time.
Newsletter image by Hannes Wolf on Unsplash