Bloomberg Tech is live from coast-to-coast. Caroline Hyde in New York, Ed Ludlow in San Francisco.
Coming up, Sam Altman says that Meta offered his employees signing bonuses as high as $100 million.
And a win for the crypto industry after a stablecoin bill passes the Senate. We discuss it with David Sacks.
Another extension of the sale on TikTok. Frank McCourt Jr. talks about his bid.
In these markets, the macro perspective is that the Fed decision is coming in a few hours and Middle East tensions. Will we see outreach between the Middle East and Iraq?
Most of the Magnificent 7 names are trading higher. Ed, you are focusing in on one particular name?
Looking at Meta, OpenAI, Scale AI, they have all been in the news a lot recently. The headline right now is because of this.
They started making these giant offers to a lot of people on our team. Like $100 million signing bonuses per year. It is crazy. I'm happy that at least so far none of our best people have decided to take them up on that.
Meta, Bloomberg has reached out to, but we haven't heard back. For more on the competition for talent, Jackie joins us.
We didn't hear the rest of that conversation with his brother. He went on to say some damning things about Meta and the fact that they lack innovation and that ultimately he is losing talent because they want to stay for the mission of OpenAI. How fierce is it out there right now?
A few things. One, it confirms that there is a new chapter in AI talent. There have been those fears but it is reaching a new level of $100 million bonuses, something we have never heard of before. It is telling of how much Meta is willing to pay to regain the edge in the AI race. Altman's comments also highlighted his confidence that it will take more than just money to bring the best and brightest to your team. If you look at the mission statement from OpenAI, they say that they think about their work for the benefit of humanity. That's really bold, large, purposeful. Meta, I looked up their mission statement earlier and it says that they are looking to accelerate the potential of future Meta products for people. Not as deep if you dig into it.
Part of the conversation that Sam Altman was having, the reasons why $100 million isn't enough to get OpenAI employees to go to Meta, what was his argument?
Part of it was the larger purpose and the fact that they are not as innovative. People want to go to a place where they feel that they are going to be able to do their best work. Some of that culture problem that we have seen with Meta, they have admitted in the past that they have trouble with middle management and just kind of the internal bureaucracy of how you get things out there. That doesn't exist to the same degree, according to Altman, at OpenAI. So when you think about the people who are going over, they will want to be able to have that assurance that their ideas are going to make it into reality.
They want to be on the winning team, they want to get to AI first, there's no point coming second. Being the business that leads the charge, those pay packages have emerged on some significant talent to the super intelligence team. From DeepMind to Sesame.
We have to remember that they are working in a new kind of organization. Meta, at the end of the day, needs to incorporate this technology into its products, which is the money maker. They have shareholders they need to appease on a quarterly basis. That kind of pressure could also affect the way the speed at which they work, how quickly they roll out these tools. Super intelligence, this is new ground for Meta, they haven't really treaded this kind of landscape before in the same way that OpenAI was founded on these principles to create a TI. You have the best and brightest going, really intelligent people, but we have seen some friction, some tension going on in its own research team inside of Meta. They are going into a place that has all of this figured out.
Jackie, thank you. OpenAI is phasing out the work it does with Scale AI, cutting ties with that company just days after the Meta investment. Rachel joins us with more. There is an important distinction happening here. OpenAI has an existing contract to Scale, but we have learned that it is going to change in the future?
Yes, OpenAI is phasing out the remaining work they are doing with Scale AI. The company branched out a lot over the last few years as data needs expanded and changed based on the kinds of models they were building. At this point, Scale represents the company putting a very small fraction of their business in, but over time they are going to be now totally phasing that out.
These two stories, yours and Jackie's, lived to the narrative. They went to the super intelligence team at Meta and took a lot of the talent from Scale AI there. Other companies start to move away from Scale AI. Have you heard this elsewhere, contracts getting undone in the future?
We are definitely hearing the companies that provide data and data services, it's a fractured and fragmented space. A lot of companies are much less known than Scale but are also worked with by these companies like OpenAI, Anthropic, Google, many others. They are seeing a surge of inbound interest and are hearing from customers some nervousness. It's possible that this tie up, they have said that they would remain as an independent business and it could make people nervous.
We should say that Scale AI declined to comment on the story as Meta declined to comment on the Sam Altman claims. The history of Scale is interesting, there is an element of manual work that they do and they rely on this army of contractors. Explain the basics. My reading of your story is that OpenAI was thinking about a relationship over the past year before a lot of this happened anyway?
Yes, Scale AI, you are absolutely correct, it started out as a business that had a large number of contractors that were paid to label data or AI models. Over time, though, the AI market has shifted quite rapidly. At this point, while it still has that legacy human data labeling business, it has expanded into a few other areas, increasingly hiring experts, experts that might work a few hours on the nights or weekends, someone with a PhD or a law degree. Instead of labeling the data directly, they are coming up with different tasks for the models to solve, something that has become increasingly important to OpenAI, but it seems that Scale, OpenAI felt the Scale was not quite up to the level it needed to support its products with this kind of work.
Rachel, appreciate that, thank you. Let's talk about the Elon Musk AI startup now. XAI is burning through $1 billion per month. Expected losses of $13 billion in 2025 because of the high costs of building those models. That's all according to sources who say that to cover the gap, his startup is trying to raise a starter for debt and equity but they have plans to spend more than half of it in the next three months alone.
Coming up, the CEO of Klarna and the White House crypto czar joins us to talk about the passage of the Genius Act. This is Bloomberg Tech.
The Amazon CEO says he expects his workforce to shrink. In an email yesterday he wrote that over the next few years Amazon expects to reduce the total corporate workforce. As the latest CEO to mourn on headcount, Amazon had 1.5 million employees at the end of March, mostly in warehouse jobs.
Another company warning about the impact of AI on the workforce is Klarna, who is racing generative AI tools. There was a change of heart on hiring for customer services. Today the startup is entering the mobile market with plans to launch a $40 per month service in the U.S. with plans to expand into the U.K. and Germany later this year. The Klarna CEO joins us now. Welcome. This was a headline I didn't expect to see. The simple question is why, why do this?
It comes back to the fact that we do believe that the future is a digital financial assistant that helps you save time and money. In this case, if we not only identify that one of our customers can save money doing this, we can bundle it together and offer it in a seamless experience by saying yes I would like a better price or offer, here it is live on my phone, that makes sense if it is a way for us to bundle more services and value to the existing customer base.
$40 a month, on the AT&T network, just want to make money on the arrangement?
We do, but most importantly it adds value to the customer base. One of my beliefs is that if we are going to have truly digital financial systems that can recognize, can wake you up in the morning and say I can save on your bills, the only thing you need to do is say yes and I will take care of the switching for you, all the paperwork, will save $10 a month or whatever it is. The point is what it means is the excess profits that had existed in utilities, banking, all of these things, will have to come down. We are predominantly stuck with our existing providers not because we love them and they are awesome, but because we hate switching and reading boring legal terms. Is this offer really better or not? The point is, we believe that the margins in these services will be significantly lower going forward. That is why also it is smart of the new entrants into the industry to come in with a very attractive offer from day one, rather than believe that we can lure people into some lengthy contract and make too much money on them over time.
New entrants already served 25 million active users in the United States, users who associate with buy now pay later. How is that narrative shifting and people starting to see you as an everything app?
Definitely with neobanking, it comes back to that financial assistant, saving time and money, don't worry about your finances. We want to help with all of that. That is kind of what banking was supposed to do, but at some point in time the banker got smart and realized you didn't have a good understanding of your own stuff and put you in some pension funds where they were charging you high fees. That has to change. We have to come back and be customer obsessed and offer the greatest value at the lowest and highest quality at the most affordable price. So, that is what we are doing and to us, it is everything related to your everyday spending. Phone is a part of your everyday spending, right, everyday costs for you. This folds neatly into that.
You, the Trump Organization, everyone is getting into phone companies. You say customer obsessed. To be customer obsessed, you have to be employee obsessed. I'm interested in the shift you have made. You came on the show before said you were all in on generative AI. I could replace myself as CEO. I don't need to add people in headcount. You have changed your mind on that?
Partially. What we said is look, in many countries you will see that in the last 10 to 20 years, doctors and nurses spend less and less time with patients and more and more time on administrative tasks. The costs of administration and health care has risen more than doctors and nurses being available to patients. What this technology enables us to do is again take the administrative burdens out, make those processed by AI, and allow doctors and nurses to be available and actually spend time on patients. It's the same for us as a company. We can use the AI technology so that when people want to speak to us and if they want to speak to a human, we want to make sure that everyone who wants to can do so, we want to make sure that those humans have less administrative tasks and can spend more time. To me, the value of human connection rises in the world of AI, it becomes more valuable and more expected that you have time for people to talk to you properly and not rushed in the customer service center, they can sit down with you and resolve it. For us, two things can be true at the same time. Use AI to automate the boring and mundane, at the same time recognizing that people will expect a human connection.
Being an everything app or platform is a goal and vision you have shared with others. X has similar ambitions, particularly in finance. Uber would also claim to be moving into those areas. Where do you see yourself sitting and who are your rivals?
It's the billion dollar question, my friend, in Silicon Valley right now. Will there be one assistant, one ring to forge them all? Or will there be multiple assistants? The answer is nobody knows. I for one need to know that our digital financial assistant needs to be the best in the world at what it does for our customers and hopefully people want to use it for that. Now, on the other hand as much as some people are saying of course there will be one, I've been around for 20 years, long enough to remember 2007 when Facebook was supposed to be where all of the internet was going to happen. That didn't actually materialize that way. So, I am still probably in the camp that there will be a few really smart assistants helping you on different topics and that you trust more on specific topics than others. But this is the billion dollar question.
We are old enough to remember that, older than Facebook. So, the IPO not happening this year? Next year?
I would say that market conditions are brighter than back in April.
As you know, I would love to say more, but I can't comment on that. At least it's great to see that markets look better.
Sebastian, thank you. Coming up, Amazon is looking to mass-produce a robotaxi. We talk to the CEO, next.
Zoox CEO says they have the capacity to turn out 10,000 robots every year, purpose-built robotaxis in Amazon Las Vegas. The Zoox CEO joins us from the new facility. 10,000 robotaxis per year at full-scale means a lot, but what is the reality? I've always known you to be very pragmatic. Give me that target for this year and next year for what will roll off the line.
The reality, really, first of all great to have you, but the reality is we are just ready and very excited, finally heading to our commercial launch later this year. From there, you have more and more robotaxis coming in. San Francisco is after that. Austin and Miami are coming up, gearing up to be basically ready. Robotaxis are purpose-built for serial production in the United States, putting that capacity in place and turning it on little by little going city by city.
10,000 is important in the terms of federal rules, right? Right now there is a limit on the number of vehicles that can be on public roads without driver controls, just like your vehicle. How are those talks going? Do you think you will get that exemption to raise scale in the coming year?
