The $10 Billion Startup Training AI to Do Your Job

0:00Travel smarter, not harder, at America's Best Value In by Sonesta, with convenient locations from coast to coast and value-packed comfort at every turn. It's a practical choice for road trips, quick getaways, and everyday travel that keeps things simple without sacrificing comfort. And when you're a Sonesta Travel Pass member, staying at America's Best Value In means earning points toward free nights, upgrades, and more every time you stay. Go to Sonesta.com to book your stay and unlock the best rates with Sonesta Travel Pass. Here today, roam tomorrow. Join now at Sinesta.com. Terms and conditions apply.

0:33If you follow markets, you know the value of long-term thinking. You plan, you diversify, you prepare for volatility. But even the best strategies can't prevent every bad day. For more than 75 years, Cincinnati Insurance has helped individuals and businesses navigate tough moments with expertise, personal attention, and independent agents who focus on relationships, not transactions. The Cincinnati insurance companies. Let them make your bad day better. Find an agent at c-i-n-f-i-n dot com. The thing about AI for business, it may not automatically fit the way your business works. At IBM, we've seen this firsthand.

1:16But by embedding AI across HR, IT, and procurement processes, we've reduced costs by millions, slash repetitive tasks, and freed thousands of hours for strategic work. Now we're helping companies get smarter by putting AI where it actually pays off, deep in the work that moves the business. Let's create smarter business, IBM. Bloomberg Audio Studios. Podcasts, radio, news. For tech workers, it's been a long season of layoffs. Meta telling staff that it will cut about 10 % of its workers in order to boost efficiency and offset its heavy spending on artificial intelligence. Block, Jack Dorsey's company, up more than 23%.

2:02It's cutting 4 ,000 employees. It's reducing its workforce by nearly half. Meta and Block, Amazon and Microsoft, they've all announced plans to downsize as they lean into AI. The dramatic moves only add to the sense that artificial intelligence could be coming for human jobs. Maybe not quite yet. Data show the labor market isn't as bad as it feels. Today, the U.S. Federal Reserve announced it would be keeping rates unchanged, citing, among other things, a stable unemployment rate, low job growth, and elevated inflation. Reporter Tom Foster has been digging into how AI could actually transform white-collar work, ever since he noticed a company called Mercore popping up on online job boards.

2:50Mercore seemed to be hiring for all sorts of roles, from filmmaker to quantum field physicist. So somebody in my background, you know, I'll see a listing that says, editor-in-chief. And I'm like, that's interesting, sure, but it's hourly. And how does that make sense? But the pay is really good. It's very good hourly pay. Tom wanted to understand what this company was all about. So he started doing some reporting for Bloomberg Businessweek. Just trying to figure out what is this thing and then realized, oh, this is the future. Mercore isn't really offering a well-paying hourly editor-in-chief role itself.

3:26But it is hoping an experienced editor or journalist will be intrigued enough to apply and teach artificial intelligence systems what only an editor-in-chief would know. Mercore is just the middleman. The training gigs are for companies like OpenAI, Anthropic, or Google. What you're doing is you're lending your expertise in your field to the AI models to train them how to do your job. Through interviews with some of Mercore's founders and contract workers it's hired, Tom has pieced together the company's business model and its ambitions to transform work itself. What the future of work looks like in this vision is humans training AI.

4:13I'm Sarah Holder, and this is The Big Take from Bloomberg News. Today on the show, the upstart company that's trying to help turn big chunks of white-collar jobs over to AI, making its founders the richest young tech entrepreneurs since Mark Zuckerberg, and providing a blueprint for a future in which we all work in the gig economy.

4:38Tom Foster says that when you answer a Mercor job listing, things can move fast. You get pretty quickly ushered into an interview and you're being interviewed by a robot. And then you're talking to this AI and the AI, you know, asks you about your job and asks smart questions. What kinds of jobs is Mercore hoping to automate or hoping to work with these white collar professionals to train AI models to do? The question is, what kinds of jobs are they not? I mean, like everything, really. The way they talk about it is that the first sort of round of AI training kind of came from the internet. And it was more or less general knowledge.

5:20And these models know a lot about all of these fields. But there's a difference between knowing a fair amount about these fields and being able to do these jobs. And that is the bridge that all of these companies are trying to cross. And to do that, it requires not just hiring thousands of doctors, lawyers, journalists, etc., but creating these exercises, these very, very specific regimented exercises that teach the models step by micro step how to do every particular aspect of that job. It's not hard to imagine how, say, an administrative assistant might train AI to do much of their work. But Mercore is trying to do that with highly specialized roles that have historically required a human touch, like medicine.

6:12So it's like, you're a doctor, here are lab reports, let's learn how to read the labs, what are the diagnoses we should make from them, what are the sort of plan of action we should put together from that. Or social work. There's nothing quite as fulfilling as when you're able to connect with someone and help them kind of progress when they're in a really difficult time. That's Tasha Kozak, a social worker Tom spoke with. She's based in Florida. She works for a giant school district there. And, you know, she really loves her work. But she, you know, a lot of the time she spends on her work is doing administrative stuff.

6:50I think AI is fantastic for helping to put together reports, to consolidate information, for data tracking. And so she figures, you know, hey, I'm going to train the AI to take away all these sort of administrative parts of my job. Tasha told us a little bit more about what she does when she clocks into her Mercore gig after school or on the weekends. To train an AI model, Tasha would, for example, provide information about a hypothetical student and teach the AI to interpret that data. For example, giving some student details, maybe some fictional medical records that I created, a story about their family situation or their current personal issue.

7:35And then the prompt itself would be to create a certain report for the student. Reports like an individualized education plan, tailor-made for the fictional student. Tasha works on a team with other Mercore workers. When I started, you had a different selection of social workers who would review the work, and then you had reviewers who would review the reviewer's work. Really just a hierarchy of different levels of review, which I think is really important because another person is going to have a different perspective and a different experience. Social work feels like such a human job, not only the interactions you're having with clients, students, but also even the paperwork and the kinds of decisions that you're making about people's futures.

8:21I'm wondering if you can talk me through how you think about training an AI to take on that role. At the end of the day, AI is a machine. And even though humans are training it, it always takes a human eye. If you're making a determination of, you know, suggesting next steps for the academic future of a child, you need that set of eyes. And you need to also communicate with the person that you're making these decisions about. In the center of social work, we say meet the person where they're at. I think it would be a very long time until any sort of AI concept or technology will be able to do that.

9:02Tasha isn't too worried about AI fully replacing social workers anytime soon. But Tom says the work she and others do for Mercore is raising more fundamental questions about the real value of a human touch. How much of our work actually requires that human skill? We've all thought the line is in one place. But as the AIs are learning these jobs, I think we're all learning the line might be in a slightly different place than we thought it was. You also spoke to a doctor of internal medicine. You spoke to a novelist, screenwriter. These are really a wide range of professions that people are bringing to MerCore.

9:42I'm wondering what is in it for them. Like, why do all these people end up working for MerCore? There are a whole bunch of reasons. It's fascinating. The writer, this is a person who's done numerous television shows, written novels, like a very successful screenwriter and fiction writer. And she, ever since the Hollywood writers' strike, she's currently sort of underemployed. Tasha Kozak, the social worker, she works 40 hours a week doing social work for the school district. She works another 20 hours working for Mercore every week. And she makes as much money in those 20 hours working for Mercore as she does in the 40 hours working in her actual full-time job.

10:23The doctor I spoke with was really fascinating. She's an internist. And her take is, I'm doing this work so that I will not be obsolete because this is coming whether we like it or not. let's learn how to be a part of it. But isn't there a limit to how much work you can get from a company like Mercore? If you train the AI, they can do the job as good or better than you. I had this thought too, as I was sort of going through it, I was like, well, what happens when they're done? We've done it. We've trained the robots. Let's close down the shop. Their take is there's just an infinite amount of this.

11:00We're nowhere near where that is in terms of what the variety of jobs that can be trained for is or the variety of tasks within those jobs. But that said, there's this whole other aspect that once you're in the Mercore system, these are short-term jobs. These are part-time jobs. There is no guarantee how many hours per week you're going to get. There's no guarantee how long this gig is going to last. There's no guarantee if this gig ends that there will be another one for you or when or that it will pay the same. So I want to talk a little bit more about the young founders of Mercore. Who are they and how did they get their start?

11:43These guys, they met in high school in Silicon Valley. There's three of them. They all come from sort of tech family backgrounds and, you know, parents are software engineers or one's father was a founder of a company, a tech company. And one of the three founders was recently sort of moved out of an executive position into a chairman position. And the other two who remain are now co-CEOs, Brendan Foody and Adarsh Hiramoth. They just turned 23. They all went to college for two years, did the typical drop out of college and start a company. In this case, with the help of Peter Thiel, they applied for one of those Thiel fellowships where you get a couple hundred thousand dollars to start, you know, a moonshot company.

12:28And their idea was not actually training AI. Their idea was, this is such a classic Silicon Valley thing. Brendan talks about the labor marketplace. This is a guy who's never had a job in his life, right? The labor marketplace is the most inefficient marketplace, you know, in history on earth. There are, you know, billions of people looking for jobs, billions of jobs out there, and no central clearinghouse. Has he heard of LinkedIn? Well, what he wants to do is like several orders of magnitude more specific and maybe invasive than LinkedIn. What he wants to do is basically, ultimately, he thinks there won't even be like real job jobs.

13:07Everything just sort of like becomes a gig, and it becomes this global marketplace of talent. Companies don't really need employees. What they need is work done. This is his idea when they founded the company. They will be the one company that has all the data on every person on the planet and their professional abilities and every single job or task or whatever that a company needs done will then be hired through this global database that this company has of all the sort of labor data in the world. They claimed that they started building this global labor marketplace, this LinkedIn on steroids, and realized in the process as they were working with some companies, what these companies need are domain experts in all these other fields as they're training for AI.

13:52And so they pivoted to that. Venture capitalists are very eager to fund AI projects. And here was one that is essentially infrastructure for the AI boom. So they started raising a whole lot of money. They're valued at$10 billion. They've raised about$500 million from the likes of General Catalyst, it's Benchmark, it's Peter Thiel, it's Larry Summers, Jack Dorsey, you know, this sort of like murderer's row of venture capitalists and other investors. What future of work are they building towards? Yeah, I mean, I'm not entirely clear on that. I mean, I tried to have these conversations about like, you know, the meaning of work and what we as humans draw from our work and, you know, the craft that we've perfected and all this.

14:39It's not really about that, right? Like, it's a very Silicon Valley way of thinking, right? Where everything is just looked at sort of in the abstract, right? And in the abstract, it's this idea that like, sure, there are probably going to be some job jobs, but like for the most part, what companies want is tasks completed. So, right, like you've got AI agents doing the jobs and you've got humans sort of overseeing them and training them and making sure they're doing it right. So in some ways, what people are doing for Mercore right now is a glimpse of the future of work. After the break, the people questioning Mercore's vision of the future and the investors who are betting on it.

15:20For many men, mental health challenges aren't recognized until they've already taken a toll. Work pressure, financial stress, changing relationships, and traditional expectations around masculinity can quietly wear men down, often without clear warning signs. In Season 3 of The Visibility Gap, Dr. Guy Winch and his guests explore how these pressures show up, how to spot them earlier, and how men can access meaningful support. Listen to the new season of The Visibility Gap, a podcast presented by Cigna Healthcare. Everyone has been there. Your team's feedback is scattered across emails, chats, and sticky notes.

15:57It's a mess. But PDF Spaces and Adobe Acrobat gives you one collaborative workspace to streamline every file and comment. So, if you need six departments to finally agree on a proposal, do that with Acrobat. Need to turn a mountain of feedback into one plan of action? Do that with Acrobat. Want to stop searching for files and finally get everyone on the same page? Do that. Do that. Do That With Acrobat. Learn more at adobe.com slash do that with acrobat. Support for the show comes from Public. Lately, it feels like there are two types of investing platforms. Some are traditional brokerages that haven't changed much in decades, and others feel less like investing and more like a game.

16:40Public is positioned differently. It's an investing platform for people who are serious about building their wealth. On Public, you can build a portfolio of stocks, options, bonds, crypto, without all the bugs or the confetti. Retirement accounts, yep. High-yield cash, yes again. They even have direct indexing. Public has modern design, powerful tools, and customer support that actually helps. Go to public.com slash market and earn an uncapped 1 % bonus when you transfer your portfolio. That's public.com slash market. Ad paid for by Public Holdings. Brokered services by Public Investing, member FINRA SIPC.

17:16Advisory services by Public Advisors, SEC Registered Advisor. Crypto services by ZeroHash. All investing involves risk of loss. See complete disclosures at public.com slash disclosures.

17:32As Mercor helps AI companies upskill their machines, Tom Foster says the company has also been dealing with some human drama. It is a controversial company, no question. And these guys know that, of course. If you go spend some time on Reddit, for instance, you will see very quickly a whole lot of people complaining about Mercore and also people, you know, quite happy about the money they're making and, you know, thinking they're building the future. The complaints have to do with, you know, you take a gig and you don't know when it's going to end or whether it will end And, you know, you have no job security.

18:04You know, what Mercore says in sort of its defense is like, hey, we were upfront about that. When you think about work as a series of tasks to be completed by task completers, these humans are just kind of machines and you treat them like machines. Right. And that's where when people run up against the difficult side of that, they get angry. There have been class action lawsuits about misclassifying workers. So this is the kind of thing we have seen with other gig marketplaces where it's about, you know, should these people be contract employees or full-time employees? Saying that everything that is expected of these people, they should be full-time employees, not contract employees.

18:43That, of course, would break the entire model, right? So another thing that's happened, there was a large data breach just a couple of weeks ago, as many as four terabytes of data. It's not clear exactly what was sort of exposed in this breach. But user data, as well as training data, seems to be the case, according to the hackers, what they posted on the internet. That is a big deal. Within a week of this coming out, there were at least five other lawsuits that came out that were about not protecting users' data. Almost immediately, Meta paused all of its work with Mercore. The other companies, it's not clear, but obviously were assessing their relationship with the company.

19:25How has the company responded to these lawsuits over worker classification? And how did they respond to the data breach? The company has essentially not responded to the data breach. They are investigating, you know, using a third party firm to help, you know, do a deep dive investigation. In terms of the lawsuits, same thing. You know, they acknowledge that the lawsuits are there. They won't say anything in particular about it. And Tom, I mean, we've talked about a few different ideas of how AI might change the economy, change the labor force. You know, there's people who think AI is going to automate away thousands of jobs, completely upend the economy for better or worse.

20:08People who think it's mostly just going to, you know, augment productivity, change what kind of jobs we do, and that the economy will ultimately be able to adapt. There are disagreements about how fast this is all going to happen. what does the data actually tell us about how AI will and is already changing work? That's the biggest question, right? Like they can say, hey, we're training the robots to do your jobs, but is it working? Can the robots do the jobs, right? Like what are they actually capable of? The fact is the AI is nowhere near ready to do these jobs competently at this point, but it is improving very quickly.

20:46What does Mercor's financial success reveal about at least the markets and investors' belief in AI's capabilities right now? And what is the contrary case? Are there detractors who are worried that this valuation is overblown and that AI is not going to be able to replace human work to this extent? Mercore is profitable, according to them. They're having no trouble raising money. It doesn't seem like there are investors who are pausing much. But that's the case with any of these AI companies. It seems to me that their experience is sort of in line with what we're seeing across the entire AI market.

21:24Is it all going to pan out to be worth that? I don't have the answer. And just zooming out from Mercore a little bit more, we've seen some high profile examples of companies justifying major layoffs with just this, with the idea that AI is going to be able to reduce the need for a large human workforce. I'm wondering how Mercore's founders might see this news. Do they see those layoffs as confirmation of their thesis? I mean, they believe that we are moving in that direction. Yes, that there will be job losses. There will be types of jobs that just go away entirely. What most people say, including some folks at Mercore, maybe we're seeing a wave of AI washing here, where people are doing rounds of layoffs for other reasons and attributing it to AI because that makes it look like they are ahead of the game or something.

22:25when in fact, maybe there were some performance issues with the company. In terms of what Mercore thinks about all of this, yes, they talk about Brendan Foody, the CEO, talks a lot about there will be job categories that go away, but also there will be new types of work that emerge. And whether that is new tasks within companies learning to train AI or whatever, or as he likes to say, you know, what this will do is free up opportunities to work on bigger and bigger problems. Because of all of this, we will now get to spend more time on the moon and go to Mars and cure cancer and, you know, do all these other large things because we are freeing up humans from doing all the drudgery.

23:13And Tom says that kind of future could be especially bright for Mercore. There's the question of like building AI agents for companies. Who manages them? How are they deployed? How are they trained? How are they maintained? How do all of the things, you know, once you've got this like workforce of AI agents out there, there has to be infrastructure for them. And Mercore and a bunch of other companies see that as an enormous opportunity for the company.

23:48This is The Big Take from Bloomberg News. I'm Sarah Holder. To get more from The Big Take and unlimited access to all of Bloomberg.com, subscribe today at Bloomberg.com slash podcast offer. If you like this episode, make sure to subscribe and review The Big Take wherever you listen to podcasts. It helps people find the show. Thanks for listening. We'll be back tomorrow.

24:19If you follow markets, you know the value of long-term thinking. You plan, you diversify, you prepare for volatility. But even the best strategies can't prevent every bad day. For more than 75 years, Cincinnati Insurance has helped individuals and businesses navigate tough moments with expertise, personal attention, and independent agents who focus on relationships, not transactions. The Cincinnati Insurance Companies. Let them make your bad day better. Find an agent at CINFIN.com. And now, The Laundry Hero. Brought to you by Grand Appliance. Oh, really? Okay, I'm on it. Uh-oh, what's up? Yeah, laundry's piled up and my washer just died.

25:06That's a bummer. Check out GrandAppliance.com. My brother got a great deal on a Maytag washer. Grand Appliance? My parents shop there. Mine too. Whole family. I'm on the site. Wow,$5.97 with next day delivery? Problem solved. Shop Grand Appliance. Appliance experts since 1930. You ever wonder how far an EV can take you on one charge? Well, most people drive about 40 miles a day, which means you can do all daily stuff no problem. Go to work, grab the kids at school, get the groceries, and still have enough charge to visit your in-laws in the next county. But they don't need to know that. And the best part, you won't have to buy gas at all.

25:48The way forward is electric. Explore EVs that fit your life at electricforall.org.