Listen to this post:
The setting: Meta’s earnings call in early August, 2026.
The speaker: Meta CEO Mark Zuckerberg.
Good afternoon everyone, and welcome to Meta Platforms’ Second Quarter 2026 Earnings Conference Call. Our remarks today will include forward-looking statements, which are based on assumptions as of today. Actual results may differ materially as a result of various factors, including those set forth in today’s earnings press release and in our quarterly report on Form 10-Q filed with the SEC. We undertake no obligation to update any forward-looking statement.
I know it’s weird that I, Mark Zuckerberg, am doing the Director of Investor Relations job, but anything is possible when this speech is made up. What follows isn’t actually me: it’s what Ben Thompson of Stratechery thinks I should say on this call.
I know that Meta and myself are facing a lot of questions about AI, particularly the amount of money we are spending on capex. Our core business is an asset-light cash generation machine, so why are we spending tens of billions of dollars on AI? To answer this question I want to give you a quick recount of our history, what I’ve learned, and why I am so confident that we are doing the right thing for our future. So let’s get to it.
A Brief History of Facebook
Facebook was, as you know, the digital representation of Harvard’s analog Face Books. What was clear from the very first day we went live was the extent to which humans are, first and foremost, interested in other humans. People would spend hours clicking around to people’s pages. To put it another way, our first algorithm was human curiosity.
What truly super-charged Facebook usage, however — and which transformed the Internet — was the feed. Now, instead of actively surfing to friends’ pages to look for an update, we showed updates to you in a single feed on your homepage.
You might remember that we got a lot of heat for this decision, including protestors outside our office in Palo Alto. The lesson we took from that, however, is one that has guided us to this day: first, the revealed preference of users, as captured by data, was that they loved the feed: engagement skyrocketed. Second, we learned to trust our own — my own — product intuition, and that conviction has served us well over the years.
Another critical moment in our early history was the shift to mobile. We didn’t get this right in the beginning — more on that in a moment — but what was quickly apparent is that more access to Facebook meant more usage of Facebook. I can’t emphasize this point enough: when humans can connect to humans, they do, and when they can do it more conveniently and in more places, they do it more often.
Finally, I would be remiss to not mention Instagram. Obviously Instagram has been a major part of our growth over the last 15 years — and, I would add, we have been a major part of Instagram’s growth. To that end, an important thing to understand about Instagram is the extent to which it has evolved. Just because we gave our users what they wanted at one particular moment in time does not mean we can afford to sit still: more bandwidth first meant more pictures in Stories, and then video in Reels. Instagram has gone from strength-to-strength precisely because it has changed as technology has changed.
My Mistakes
We — I — haven’t done everything perfectly. We’ve taken our arrows through the years for lots of things that frankly aren’t our fault, but are rather the reality of being the primary communications platform for all of humanity, and humanity is flawed. I’m proud of the efforts we have made to ameliorate humanity’s worst impulses while enabling some of our best tendencies, including that desire to connect.
Rather, my mistake is itself a very human one: for many years I have resisted embracing what Facebook — now Meta — is, and spent too much time trying to emulate some of the tech titans who came before me. Specifically, I have been obsessed with becoming a platform.
The first manifestation of this error was the initial shift to mobile I referenced above. When Facebook was primarily a browser app I invested heavily in trying to build a platform, with things like Facebook Games, payments, etc. We had some success there — some of you on this call might have played Farmville back in the day — but when mobile came along we mistakenly tried to hold onto web technologies that supported my vision, and were years too late in investing in a truly native smartphone experience.
The reality — and this is hard for me to admit — is that Apple saved us from my mistaken obsession. Mobile Made Facebook Just an App, and that was Great News. Instead of diminishing the Facebook experience so that we could feature third-party developers, we had to cede that space to Apple and put our own content front-and-center. It turns out that was what people wanted the most; in fact, they wanted it so much that they willingly scrolled through and clicked on the most compelling ad units ever. And make no mistake, we paid back our debt: Facebook built the App Store just as much as Apple did.
My second error was Reality Labs. While in recent years I have framed our acquisition of Oculus and virtual reality as a necessary response to Apple’s attempt to handicap our business, the truth is that I invested twelve figures into this technology because I thought it was cool, and yes, because I wanted to own a platform. I do think we’ve made compelling strides in this area — and we’ve created technology that is going to matter in the long run — but I now recognize that part of the reason I am delivering this mea culpa right now is because I burned a lot of credibility with investors with all of the losses Reality Labs has endured with very little to show for it.
My third error was not in trying to make Facebook something it was not, but rather failing to appreciate what it had become. While I was thinking about platforms, I took it for granted that connection was enough for the core business; in fact, Facebook had evolved into entertainment, at least in its public-facing forms (I will take credit for the acquisition of WhatsApp and realizing that Messaging Was Mobile’s Killer App). This was an insight that TikTok figured out first, and it was a blindspot for me.
The Ad Blindspot
What I’ve come to realize is that all of these mistakes are symptoms of what has been my biggest failing as CEO: all of you on this call have appreciated our ad business more than I have. I’ve been very blessed as CEO to have excellent co-workers who have over the years developed the world’s best digital ad business, while I frankly haven’t taken as much interest as I should have.
My failure to appreciate our ad business is another lens through which to examine my mistakes:
- Building a platform is antithetical to building an ad business. A platform’s goal is to feature third-parties; an advertiser’s goal is to capture attention for itself.
- Investing in an entirely new technology, including developing hardware, fundamentally limits our addressable market; an advertiser’s goal is to maximize its market size.
- Entertainment is the best possible category for an advertiser to own: people willingly give entertainment their attention, which is exactly what an advertiser wants to sell.
This neglect as CEO left us badly exposed in our disputes with Apple. I firmly believe that Apple’s characterization of digital advertising was unfair, dishonest, and self-serving. What I failed to do, not just in that bruising battle but in the years leading up to it, was make the affirmative case for ads generally, and Meta ads in particular.
It’s easy to see how the Internet has made it possible for an entirely new category of entrepreneurs to create products that uniquely serve the tremendous capacity of humans to manufacture an infinite array of desires, growing the economy to the benefit of everyone; what’s harder to appreciate — in part because I haven’t made the case — is that the only way to connect those creators to the consumers who love them is digital advertising. We don’t serve ads like Google — or Apple in the App Store, or Amazon on Amazon.com — that in many respects function as a tax on search; we show people products they never knew existed, but that immediately generate desire and, ultimately, happiness. In short, I believe that we are a force for good in the world, not just because we connect people to each other, but because we connect entrepreneurs with customers in a way no one else does.
Why AI Matters
Forgive the long preamble, but this is necessary context for me to properly explain why AI is so important to Meta, and why I am making the right choice to invest so heavily in both talent and infrastructure.
First, when investors compliment our asset-light business, what they are complimenting is the fact that our business is purely digital. Everything digital, however, is firmly within AI’s cross-hairs. It may seem odd to begin my AI pitch by highlighting terminal value risk, but today is about honesty: every single digital company on earth faces an existential threat from AI, and we are no exception. Meta must invest in AI because a failure to do so would cost us far more in the fullness of time, particularly now that we’ve seen the very real risks entailed in depending on a third-party.
Second, AI makes our business better — and by “our business”, I mean ads. AI is more than LLMs: it is machine learning, and we have been using machine learning to improve our ads business for years. More recently, we have developed GPU-dependent algorithms that have significantly improved our ability to not just target ads but also recommend content, which keeps people entertained longer, which lets us serve them more ads. And, looking forward, LLMs themselves will transform advertising, not just by generating copy and images, but by predicting the ads and content that people want to see. Every single one of these improvements goes directly to our top line — and remember, because advertising enables us to offer our products for free, the capacity to increase our top line is unbounded by price elasticity.
Third, the single most important indicator that our business is on the verge of a step-change in growth is when we dramatically increase inventory. This is something investors regularly get wrong: back when we added Stories, investors panicked about falling prices-per-ad without realizing we were increasing inventory we could grow into. Five years later, investors made the exact same mistake with Reels. Those were the two best opportunities to buy Meta stock — or any stock, really — in history. We are facing an even larger opportunity over the next several years. AI makes every pixel monetizable, which means we are looking at the largest inventory expansion ever. Yes, it will take a few years to realize this opportunity, but the technology is there.
More importantly, what I’ve come to realize as I’ve embraced our status as an entertainment provider and ad purveyor is that — our nature as a digital business notwithstanding — we are remarkably well-placed to thrive in an AI era. Remember what we learned about humans: they are obsessed with other humans, and they want to connect with them; that obsession and desire are only going to increase as we interact more and more with AI. AI is going to make our properties more essential, not less.
Moreover — and here I must issue one more mea culpa — AI is a productivity tool, but productivity is not the end-all-be-all of the human experience. I have talked over the last year about building superintelligence that helps you get things done, but that’s a business story. What we can uniquely do is give people the experiences they want — from connection to entertainment to shopping — when they are off the clock. The fact that we are investing in AI but not selling solutions to businesses is actually one of our biggest advantages.
Oh, and by the way, AI might actually lead to new hardware paradigms. I admit I was wrong to spend so much time on virtual reality, but that did lay the groundwork for a unique opportunity to develop devices that make much more sense in a world where we want to access AI everywhere, not just on a phone in our pocket.
The Compute Hurdle
I know that many of you on this call have doubted my investment decisions before — and I understand the consternation about Reality Labs in particular. However, keep in mind that when our stock dipped in 2022, one of the big reasons was because of our aggressive capex spending, which went primarily to GPUs; ChatGPT came out a month later, and that decision to spend heavily with Nvidia looked incredibly prescient in hindsight.
That prescience, however, pales in comparison to the payoff that will accrue to anyone with the foresight to build data centers and buy compute over the last several years, and for years into the future. We don’t have the luxury of waiting until the future is invented and then investing; we need to invest now, especially when the opportunity in front of us — with ads specifically — is so apparent.
That noted, we are in a truly unique time, when there is a real market for selling compute on the spot market. To that end, we are going to sell access to a portion of our compute infrastructure on a short-term basis, with the ability to claw back that compute at any time. This will accomplish two important things: first, the proceeds from these rentals will fund an even larger build-out going forward. We will need this capacity in the future. Second, rental prices will provide a hurdle rate that will focus and discipline our decision-making.
Let me expound on this point, because it brings this entire opening statement together: I now realize that my obsession with platforms and productivity has frequently led us astray, and that I have given insufficient appreciation to our advertising business and failed to embrace the reality of what Meta is. I truly believe we have compelling reasons to invest in AI — arguably the most compelling reasons — and the fact that the market doesn’t agree is my failure.
To that end, making our compute available for rent means that we can only take it back if we can make more money on it ourselves; the only way we can do that is by leaning into what we are good at, not what I have spent too long wanting us to be. To put it another way, our best product decisions have been intuition validated by data and revealed preference; that’s how we’re going to approach AI. We will build, because we must, but we will let the market decide who gets to use it: I’m confident my newfound religion on ads will result in all of that compute being used by us to make more money than we can ever make as a permanent cloud provider.
We are not out here to make chatbots or compete with OpenAI and Anthropic; they can fight for work and productivity and charging subscriptions and replacing humans. Our goal is to celebrate humans, to connect them, to entertain them, and to enable commerce among them. We need compute to do this at scale, and I know it will pay off. My commitment to you is that we will structure our business so we have no choice but to do just that. We’ve done it before, and we will do it again.
And with that, over to Susan.