A (0:00)

Welcome back to the rundown for another weekend deep dive. Today we are talking about OpenAI. OpenAI is going through a rough stretch these days. Now, this company was once seen as a surefire success of the AI era, but now people are starting to ask some tough questions. There are reports of the company missing internal growth projections. There seems to be disagreement between the CEO and cfo, all while competitors like Anthropic and Google are eating into their market share. So in today's episode, we're going to talk about how OpenAI became the center of the AI universe, why the cracks are starting to show, and what happens to the rest of the tech industry and the economy if OpenAI stumbles. We got a great one for you today. Let's dive in. Before we get into the drama and challenges that OpenAI is facing right now, let's first give Sam Altman and OpenAI their flowers, because what they've built in just three and a half years is insane. ChatGPT launched back on November 30, 2022, and immediately saw widespread adoption. In fact, Chat GPT hit 100 million users in just two months. For context, it took TikTok nine months to get there. And Instagram, it took two and a half years. You fast forward to today, and ChatGPT now has more than 920 million weekly users. And that massive user base has translated into an explosion in revenue. In 2023, OpenAI did $2 billion in revenue for the full year. Now they're doing about $2 billion in revenue every month. And that's why investors and venture capitalists have been piling money into the company to get a stake in what could be a once in a generation company. OpenAI just closed a record $122 billion funding round, valuing the company at $852 billion. So that makes OpenAI worth more than JP Morgan, more than Eli Lilly, more than ExxonMobil. And it all happened in under four years. And, you know, as OpenAI grew and more and more people started using ChatGPT, OpenAI needed more and more compute. They needed access to Nvidia chips and data centers and power. So Sam Altman went on a deal making spree, signing data center deals with Oracle and Core Weave and buying chips from broadcom, Nvidia and AMD. In total, OpenAI has committed to spending $600 billion on AI infrastructure across these deals. And, you know, every time these deals would be announced, the stock price of that company would surge. It happened with Oracle, it happened with amd, it happened With Broadcom, all these companies saw billions of dollars added to their valuation off the OpenAI announcement. So for most of the last couple years, OpenAI effectively made itself the center of gravity for the entire AI economy. They became the company that everybody else's stock price became tied to. And you know, that's great when things are going well, but it also means that if OpenAI stumbles, a lot of these companies are, are going to feel it. And recently there have been some stumbles at OpenAI. So let's talk about the cracks that are starting to show. For the past three years, OpenAI has been the undisputed king of AI. ChatGPT has basically become synonymous with artificial intelligence. Like, if you ask your parents to name an AI product, they're probably going to say ChatGPT. But that's starting to change now with competition from Google and Anthropic. Let's start with the rise of Anthropic. This is the company behind Claude. You know, for a while, Anthropic was seen as a scrappy underdog. The company was founded by former OpenAI employees back in 2021 who left OpenAI over disagreements about safety. And for the longest time, Cloud was mostly known in, like, the techie circles. The average person had no idea what Claude was. But while most people weren't paying attention, Anthropic was quietly winning over the enterprise market. Companies were choosing Claude over Chat GPT for, for coding because, frankly, it was just better than ChatGPT. Anthropic's API business was growing, and today they have over a thousand enterprise customers spending more than a million dollars a year. And, you know, this year, Anthropic also broke through with the consumers too. I think it was the launch of the Opus 4.5 model that really tipped the scales. That model was absolutely incredible. On top of that, Anthropic also started releasing tools like Claude Code and Claude coworkers that regular people could use and not just developers. So that resulted in an explosion in adoption. I mean, I personally went from using Claude like 20% of the time last year to now using it like 90% of the time. And, you know, all of that is showing up in their revenues. Anthropic's annualized revenue just hit $30 billion as of April. To put it in perspective, how fast their revenues have grown, their run rate was $1 billion in January of last year. So they've 30x their revenue in just 15 months. In fact, Anthropic may have overtaken OpenAI in revenue, which was sitting around $25 billion as of the most recent reporting. Now, OpenAI disputes how Anthropic counts as revenue, saying that it's not apples to apples comparison. But the fact that we're even having this conversation tells you how much the landscape has shifted and investors are paying attention. Anthropic is now reportedly seeking investments at a $900 billion valuation, which would put them at a higher valuation than Open AI. So Anthropic has gone from a scrappy underdog to arguably OpenAI's biggest threat in less than a year. Meanwhile, you also have Google's Gemini, which has been quietly growing its user base from 350 million last year to more than 750 million monthly users as of February of this year. See, part of what makes Gemini a tough competitor for Chat GPT is that Google can afford to subsidize it because of their existing business, making $400 billion in revenue a year. On top of that, Google already has a billion plus existing users and that they can easily introduce Gemini to. And the key is that Google has multiple ways to monetize Gemini, whether it's through ads, whether it's through subscriptions, whether it's through their cloud platform. So Google has become a legit threat to OpenAI. And you know, the competitive threats from Google and Anthropic is starting to impact ChatGPT's growth. According to the Verge, monthly user growth has cooled off significantly. ChatGPT's monthly active users jumped 168% back in January, but that slowed to just 78% by April. And if you look at ChatGPT app downloads, they were up only 14% year over year over the past few months. And then there was a report from the Wall Street Journal that dropped this week that said that OpenAI missed its internal goals of hitting 1 billion weekly active users by the end of last year. Now, OpenAI still hasn't announced that they've hit that 1 billion user milestone. So that probably tells you everything you need to know. And look, OpenAI is still the biggest player in AI, but that growth is slowing and the gap is closing. And that matters a lot when you're trying to justify an $852 billion valuation and the $600 billion in spending commitments. So that brings me to the bigger question. If growth is slowing, does OpenAI have a business model problem? Let's talk about it. OpenAI is closing in on 1 billion weekly users, but only about 5% of those users are actually paying to use Chat GPT. So that means that 95% of people using ChatGPT every day or are using it for free and generating zero dollars in revenue for OpenAI. In fact, every single one of those free users is costing OpenAI money because they're using computing power every time they use the service. So that's the challenge that OpenAI faces on the consumer side and then on the enterprise side. I mean, Anthropic has been the dominant player. And it's not just because Anthropic's models are better at coding compared to ChatGPT, which most people say they are. It's also because Anthropic's models are more widely available. See, until this week, OpenAI's models were exclusively available only through Microsoft's cloud platform, Azure. And that's because as part of Microsoft's original investment in OpenAI, back in 2019, there was this exclusive partnership. But, you know, a lot of big enterprises run their businesses on Amazon Web Services. So if you're a Fortune 500 company running on AWS and you wanted to use OpenAI's models, you either had to go through Microsoft or you just went with Anthropic models, which have been available on AWS for a while now. So I think that helped Anthropic get a leg up when it comes to the enterprise. So OpenAI's business is getting hit from both ends. Right now, their consumer business only has about 5% of users paying, and their enterprise business was being held back by a cloud exclusivity deal. In this Wall Street Journal article from this past week that I referenced earlier, it also mentioned that OpenAI missed its internal revenue targets for the last year, and that has reportedly spooked some people inside the company. OpenAI CFO Sarah Fryer has told People that she's worried the company won't be able to pay for its computing contracts if revenue doesn't grow fast enough. Remember, Sam Altman has committed roughly $600 billion in future spending commitments. But look, OpenAI is trying to expand their business model to grow revenues again. You know, this past week, OpenAI and Microsoft restructured their deal so that OpenAI can now offer its products on any cloud, not just Azure. And literally the very next day, after modifying this deal, OpenAI announced that its models would be available on Amazon Web Services through Amazon Bedrock. So this is OpenAI finally opening the door to to all those AWS customers. The other big move that OpenAI is trying is ads. OpenAI started testing ads on the free version of ChatGPT back in February, and it's been a pretty good success. So far, according to a report from the Information, the Chat GPT ad pilot hit $100 million in annualized revenue within six weeks. And by 2030, OpenAI is projecting that ads will be their single biggest revenue source, at over $100 billion in revenue and and accounting for about 36% of total sales. You know, I personally, I feel like OpenAI should have launched ads earlier, because the reality is the best way to monetize a consumer business on the Internet is through ads. I mean, just look at Google and Facebook. The truth is, most users just don't want to pay for stuff on the Internet. If OpenAI started testing ads a year or two ago, they would be so much further ahead. But at least it's better late than never. Now, OpenAI has also dipped its toes into shopping. They launched a feature called Instant Checkout last fall where you could purchase products directly inside ChatGPT. Companies like Etsy, Walmart and Shopify all signed up, and OpenAI said they would take a small cut on the sales generated through their platform. Well, after a few weeks of testing, this thing never really caught on. The shopping experience wasn't great. And Walmart actually found that conversion rates for purchases made directly in ChatGPT were three times lower than when users were sent to Walmart's own website directly. So now OpenAI is already pivoting away from the Instant checkout and moving to a model where retailers build dedicated apps inside ChatGPT that redirects users to the retailer's website for the actual purchase. So to sum it all up, OpenAI is trying to fix their business model and reignite revenue growth by expanding to new clouds and adding advertising and getting into E commerce. I think some of these moves are smart and probably overdue, like the US Expansion. Others are still unproven, like the ads pivot. And some have already stumbled, like the shopping thing. The question, though, is whether all of this happens fast enough to keep up with the $600 billion in spending commitments. And listen, this is not just an OpenAI problem anymore, because if OpenAI can't pay those bills, a whole lot of other companies are going to feel the pain. So let's talk about the domino effects of OpenAI stumbling. So here's the thing that makes the OpenAI story worth paying attention to. As an investor, OpenAI has made itself so central to the entire AI economy that when it sneezes, the. The whole tech sector catches a cold. And we literally saw this happen in real time this past week. On Tuesday morning, after the Wall Street Journal dropped their report about OpenAI missing their targets. Shares of companies like Oracle, Broadcom, AMD and Core weave all fell 3 to 5%. So billions of dollars in market value were wiped out, all because of one report about one private company that doesn't even release its financials to the public. And that should tell you how deeply OpenAI is woven by into the financial fabric of the tech industry right now. I mean, think about the relationships here, right? Oracle has a $300 billion five year deal to supply computing power to OpenAI. Microsoft has invested $13 billion into OpenAI. Amazon just invested $50 billion into OpenAI. SoftBank put in 30 billion. Nvidia put in 30 billion. And AMD has a deal reportedly worth $90 billion. And those are just some of the deals that I listed. So if OpenAI has a slowdown and can't pay its bills, well, it's not just OpenAI's problem, it's Oracle's problem. It's Core Weaves problem, it's AMD's problem. There could be ripple effects not just across the tech sector, but the entire economy. See, all this spending on AI is actually boosting US GDP. The US economy grew 2% last quarter, and a big chunk of that was driven by AI related business investments. You know, all this spending on data centers is creating jobs for other industries like construction and increasing demands for plumbers and electricians and many, many more types of workers. So it kind of feels like the entire economy right now, and the stock market for that matter, is being propped up by AI. And that's why OpenAI's struggles have come into such scrutiny and caught the attention of everyone. Because the reality is millions of portfolios and retirement accounts are counting on OpenAI's success. Maybe OpenAI's growth slowing down is the first sign that the AI bubble is on the verge of popping. Or it's possible that we're blowing all of this out of proportion. So what's my take here? Well, I might be in the minority here, but I kind of feel like this whole OpenAI thing is a bit overblown. Like, sure, they might be struggling right now and have to scale back some spending and renegotiate some deals, but this company just raised $122 billion and has 900 million plus weekly users. They're not going anywhere. Okay. And look, even if OpenAI does disappear tomorrow, I think the overall ecosystem would be just fine. Sure, some companies would get hit harder than others like Oracle and Core, we've come to mind. But the demand for AI compute right now is so massive that every other major player would be lining up to buy those chips and fill that void left by OpenAI. In fact, the four big tech hyperscalers, which includes Amazon, Google, Microsoft and Metta, just reported earnings this past week and they spent a combined $130 billion on capital expenditure in the first quarter alone. And these four companies combined plan to spend over 6, 700 billion dollars in CapEx for the full year. So the AI spending boom is way bigger than just OpenAI at this point. Now, where I do think there is a legitimate risk is investor sentiment. Even if OpenAI's problems don't actually break the AI economy, they could break investor confidence in the AI trade. We've seen this play out before. After the dot com bubble, one or two high profile failures cause investors to panic and pull money out of the entire sector and even from companies that were doing well. So if OpenAI keeps missing targets and that becomes a bigger story, you could see some nervousness in names like Oracle, Nvidia, Broadcom and other AI infrastructure companies. But personally, I think the most likely outcome is that OpenAI will just tighten up their spending and their ad business will scale faster than most people expect. On top of that, this AWB partnership will probably open up a whole new enterprise market as well. Now, OpenAI might not become a generational once in a lifetime company that some people expect. In fact, if you invest in OpenAI right now, you might not even double your money ever. But I think there's still going to be a major player in AI now. I wouldn't be surprised though, if Sam Altman gets replaced as CEO at some point. You can make the case that he got way too distracted with too many side quests over the last couple years. OpenAI was launching browsers and working on hardware and trying to build a shopping platform that flopped. Meanwhile, they didn't take into account the competitive threat from Google and Anthropic seriously enough. So he kind of let them catch up and in Anthropic case, maybe even overtake them. So Maybe all that OpenAI really needs is a new CEO focused purely on execution. Because remember, this company is trying to IPO sometime this year and you know, who knows if that's still going to happen because their CFO reportedly doesn't think they're ready. And honestly, based on all the reporting, I think she might be right. But selfishly, for content reasons, I do hope that OpenAI does move forward with their IPO sometime this year. Well, all right guys, that's it for today's. Weekend deep dive. Hope you guys enjoyed that one. You know this is definitely a story we're going to be following closely, especially if the IPO actually happens this year. So make sure you guys are subscribed to the podcast so you don't miss the follow up episode. Let me know in the comments on what you thought about today's episode. Do you think this OpenAI panic is overblown or do you think there's some real trouble ahead? Also, would you invest in OpenAI right now at an 850 billion dollar valuation? Drop your thoughts on Spotify and YouTube. Also while you're at it and you have like 5 extra seconds, consider giving us a 5 star rating on Apple, Spotify wherever you listen to your podcast now. All that engagement really does help us out and it helps other people find the show. Thank you guys again for listening, watching and commenting. Shout out to Mike and Connor for all the work behind the scenes and we'll see you guys back here tomorrow.