Uber Is Making A Push In Data Labeling After Scale AI’s Deal With Meta

Here's a story that might surprise you. You know Uber, right? The company that gets you from point A to point B, or delivers your dinner when you're too lazy to cook? Well, they're making a major play in a completely different game - and it's all because of some serious drama in the AI world.

Let me set the scene for you. Last week, the tech world got hit with a bombshell. Scale AI, one of the biggest names in training artificial intelligence, made a deal that nobody saw coming. Meta - you know, Facebook's parent company - swooped in and grabbed a 49% stake in Scale. Just like that.

Now, here's where things get interesting. That deal sent shockwaves through the industry faster than you can say "ChatGPT." Suddenly, Scale's biggest clients were running for the exits. OpenAI, the company behind ChatGPT, had already been pulling back from Scale for months. Google started planning their own escape route. It was like watching a game of musical chairs, but with billion-dollar AI companies.

Enter Uber, Stage Left

And that's where our story really begins. While everyone was scrambling, Uber saw an opportunity. You see, since last November, the ride-hailing giant has been quietly running something called Uber AI Solutions. It's their data-labeling platform, and it's all about training AI models for big corporate clients.

Think about it this way: you know how Uber connects drivers with riders? Well, now they're connecting human workers with AI training tasks. It's the same platform concept, just applied to a completely different problem.

Megha Yethadka runs this operation. She's been with Uber for ten years, so she knows the company inside and out. When I think about what she told Forbes, it makes perfect sense: "For Uber, our core has always been being the platform of choice for flexible on-demand work."

She's absolutely right. Uber drivers are contractors who pick up passengers and deliver food all over the world. Now, Uber is taking that same model and applying it to digital tasks. Instead of driving people around, these workers are helping train the next generation of AI.

What Uber Is Really Offering

So what exactly is Uber bringing to the table? Well, they just announced some pretty big updates that show they're serious about this business.

First up, they're offering ready-to-use datasets. We're talking audio, video, images, and text - basically all the raw material that companies need to train their AI models. It's like having a well-stocked pantry when you want to cook a fancy meal.

But here's the really clever part. Uber isn't just providing the data - they're licensing out the same platforms they use internally. Think about it: Uber has been managing massive networks of contractors for over a decade. They know how to organize people, assign tasks, and maintain quality control at scale. Now they're letting other companies use those same tools.

And they're not stopping there. Beyond just training models, Uber is now helping clients develop AI agents. These are the smart assistants that can actually take actions for users, like handling customer support or managing schedules.

Oh, and here's a fun tidbit: the service used to be called "Uber Scaled Solutions." But recently, they dropped "Scaled" and went with "AI" instead. Yethadka insists it had nothing to do with avoiding confusion with Scale AI, their competitor. She says they just wanted to make it clearer what the unit actually does. Sure, Megha. We believe you.

The Automation Advantage

Here's where Uber thinks they can really beat the competition. They're working on automating more of the setup process for these AI training projects.

Picture this: instead of spending weeks setting up a project manually, clients can just describe what they need in plain English. The platform then automatically figures out how to assign tasks, set up workflows, and maintain quality control. It's like having a really smart project manager that never sleeps.

The goal is to get human workers involved faster, instead of wasting time on all the boring setup work. And when you think about it, this makes total sense. Uber has spent years perfecting the art of quickly matching supply with demand. Now they're applying that same expertise to AI training.

Going Global

The numbers are pretty impressive too. Uber AI Solutions is now available in more than 30 countries. That's a huge jump from the five markets they started with last November, which included the US, Canada, and India.

Since the beginning of this year, they've doubled the number of clickworkers on their platform. Yethadka wouldn't say exactly how many people are in their network, but she did mention there are "tens of thousands" working in each topic area. We're talking about specialists in STEM, coding, law - you name it.

And here's something that might surprise you: the most engaged workers are putting in about 3 to 4 hours a day on these tasks. The pay ranges from $20 to $200 per hour, depending on how complex the work is. That's not pocket change - for many people, this could be a decent side hustle or even a full-time gig.

The unit already has more than 50 corporate customers. Some big names too, like Aurora, the self-driving car company, and Niantic, the folks who made Pokemon Go and recently pivoted to enterprise AI.

The Scale Shakeup

Now, let's talk about why this timing is so perfect for Uber. Scale's deal with Meta has thrown the entire data labeling industry into chaos. As part of the deal, Scale's CEO Alex Wang is heading over to Meta to run something called the Superintelligence Lab. It's Meta's big bet to compete with OpenAI, Anthropic, and Google in the AI arms race.

Yethadka puts it perfectly: "A number of companies are, of course, looking to revisit their partner strategy for data." Translation: Scale's clients are looking for new vendors who are "neutral and impartial."

And boy, are there companies ready to pounce. Smaller rivals like Mercor, Turing, and Invisible Technologies are all scrambling to grab Scale's former clients. But here's the thing - Uber has something most of these companies don't: size and staying power.

The Big Company Advantage

Yethadka makes a compelling argument here. "A lot of companies in this space are a lot smaller, VC-funding dependent," she says. And she's got a point.

Uber is worth $175 billion and brought in almost $44 billion in revenue last year. That's not startup money - that's "we're not going anywhere" money. While other companies in this space might be worried about their next funding round, Uber can play the long game.

Plus, there's something to be said for experience. While many of their competitors are essentially service providers, Uber has been building and shipping products for over a decade. They know how to collaborate with customers because they've been doing it forever.

"We have been a product company and an operations company, and have done this for a living ourselves," Yethadka explains. It's like the difference between a consultant who gives advice and a contractor who's actually swung a hammer.

The Competition Speaks Up

Even Scale noticed what Uber was up to before the Meta deal went down. Earlier this year, Xiaote Zhu from Scale's Outlier platform told Forbes: "This space is full of opportunities. I think more people are seeing the value of the work we're doing here, which is why even a business like Uber will want to try their hand at the same space."

That's a pretty diplomatic way of saying "welcome to our neighborhood."

But not everyone thinks Uber's guaranteed to win. Brendan Foody, who runs the $2 billion company Mercor, makes a good point: "Data annotation is transitioning towards higher and higher-skilled work. Uber's success will depend on how effectively they build this high-skilled talent network."

He's right. This isn't just about having lots of workers - it's about having the right workers. The AI training tasks of today require people who really know their stuff, not just anyone who can click a mouse.

The Baggage Question

Now, let's address the elephant in the room. Uber hasn't exactly had a smooth ride over the years. They've dealt with regulatory battles, controversies about how they treat their drivers, and all sorts of public relations headaches.

Does this hurt their chances in the AI training business? Yethadka doesn't think so. She says customers haven't been bothered by Uber's past, and the company is committed to "doing the right thing" when it comes to data security and confidentiality.

"And that continues to be applied to this new line of business as well," she adds. Time will tell if customers buy that argument.

The Bigger Picture

So what does all this mean? Well, we're watching the AI industry reshape itself in real time. Scale's deal with Meta has created a massive opportunity, and companies like Uber are rushing to fill the gap.

The question is whether Uber can successfully translate their ride-hailing expertise into AI training dominance. They've got the resources, the platform experience, and the global reach. But they're also entering a highly competitive market where the rules are still being written.

What's certain is that the AI training industry just got a lot more interesting. When a company known for cars and food delivery decides to compete with specialized AI firms, you know the stakes are high.

The next few months will be crucial. Will Uber's bet pay off? Can they convince Scale's former clients to switch sides? And most importantly, can they build the high-skilled workforce needed to stay competitive in this rapidly evolving market?

One thing's for sure - the ride-hailing giant isn't content to just drive you around anymore. They want to help drive the future of artificial intelligence. And given their track record of disrupting industries, that's a ride worth watching.