--- title: "Investment of Up to $200 Billion in AI Development Revealed" url: https://newscentral.site/investment-of-up-to-200-billion-in-ai-development-revealed/ language: en publisher: "News Central Site" section: "Business" published: 2026-10-20T07:30:00.000Z updated: 2026-10-20T09:57:51.661Z id: 4be3a526-56a3-4846-89b6-17979c8a4a63 source: "Fortune https://fortune.com/2026/10/20/zuckerberg-apple-lesson-meta-100-billion-ai-bet/" attribution: "Link to https://newscentral.site/investment-of-up-to-200-billion-in-ai-development-revealed/ and name News Central Site when you quote or summarize this story." --- # Investment of Up to $200 Billion in AI Development Revealed Meta's Mark Zuckerberg candidly admitted to greenlighting the AI buildout with a budget of up to $200 billion despite uncertainty about its success. Mark Zuckerberg has revealed that he took a massive risk in investing billions of dollars into Meta's artificial intelligence (AI) development before there was any concrete evidence it would pay off. In a recent podcast episode, Zuckerberg candidly admitted to greenlighting the AI buildout with a budget of up to $200 billion, despite uncertainty about its success. He recalled discussing the project with an unnamed person, asking if they were certain it would work. This level of investment is unprecedented in the tech industry, and raises questions about the risks involved. Decisions of such magnitude are often made without concrete proof points, according to Alexandr Wang, Meta's chief AI officer. He explained that large-scale investments like this are typically made with a long-term vision in mind, rather than relying on short-term gains or immediate results. This approach is at odds with the Lean Startup philosophy, which emphasizes rapid experimentation and iteration. Meta has already committed a significant portion of its capital expenditures for the year to AI infrastructure, with a total budget of up to $145 billion. The company's investment in AI is focused on creating a "full-stack operation, where it develops and builds all aspects of its technology stack, from data centers to low-level infrastructure. For years, Mark Zuckerberg has been working around the limitations imposed by Apple and Google on their respective smartphone platforms. Facebook's mobile journey was marked by friction with these platform owners, some of which was self-inflicted. The company pushed updates to its Android app outside the Google Play store in 2013, prompting a ban from Google. This move highlighted Facebook's desire to operate independently and not be bound by the rules set by others. Apple also took issue with one of Facebook's apps, shutting down Research in 2019 after it was discovered that the platform paid users up to $20 per month for access to their phone data. The app had been accused of invading users' privacy. The biggest blow came in 2021 when Apple introduced a feature allowing iPhone users to block apps from tracking them, a change that Meta estimated would cost it around $10 billion in 2022 alone. Zuckerberg has cited his experiences with these platform owners as key learning moments in his approach to AI development. Meta's approach to AI development is set against the backdrop of significant costs associated with building and maintaining its own data centers. According to Mark Zuckerberg, a one-gigawatt AI data center can cost anywhere between $50 billion to $60 billion to construct and around $15 billion annually to operate. To mitigate these expenses, Meta has been considering diversifying its business model by launching a cloud service that would allow the company to rent out its data centers to external customers. This move could potentially generate significant revenue for the company. However, this plan is still in its infancy, with Meta's Hyperion campus in Louisiana designed to scale up to five gigawatts of computing power. If successful, it could provide a substantial hedge against the costs associated with building and maintaining its own data centers. Meta's ambitions extend beyond simply providing raw computing capacity. The company wants to offer a full-stack solution that includes not just the hardware but also the software necessary for AI applications to function seamlessly. This approach is designed to drive the development of Muse, Meta's personal AI agent. Meta's AI agent, Muse, is being used in increasingly sophisticated ways by its founder and CEO, Mark Zuckerberg. For instance, he uses it to source rare teas from specialty vendors. Moreover, he has even set up an Instacart account specifically for Muse, which now orders ingredients for weekend baking projects with his young daughter nearly every week. This shift towards agent-driven commerce could have significant implications for Meta's business model. Advertising currently accounts for a substantial majority of the company's revenue, approximately 98% last quarter. Furthermore, eMarketer projects that Meta will surpass Google in worldwide ad revenue this year, reaching $243.46 billion. However, if AI agents like Muse become prevalent, they could potentially disrupt the advertising-driven business model. Agents would compare prices, check delivery dates, read reviews, and ultimately make purchase decisions, rendering ads largely redundant. This raises questions about the role of advertising in commerce. Collison has suggested that businesses that rely on overcharging consumers might see AI agents as a threat, and therefore resist integrating them into their operations. However, Zuckerberg counters that most businesses are eager to work with agents, arguing that they could ultimately drive overall growth in commerce. He points out that many companies are already seeking deeper integration with Muse, which he believes will lead to an increase in the total addressable market for all of commerce. Meta's business model expansion plans have been revealed by Mark Zuckerberg, who sees potential in subscription services and new revenue streams beyond advertising. The company is already testing paid plans for its Muse AI platform, which offers free use with optional upgrades costing $20 or $100 per month for more intensive users. Zuckerberg has also hinted at a small commission on savings or earnings generated by Meta's business agents, a feature he likened to the ad system in terms of revenue generation. However, one key question remains unanswered: whether business agents will cannibalize ad impressions, which currently account for nearly all of Meta's revenue. For now, investors seem unfazed by this uncertainty, with Bank of America analyst Justin Post noting that Meta is back on offense" with its AI investments, singling out rising AI spending as the main concern rather than the company's emerging business strategies. --- Source: [Fortune](https://fortune.com/2026/10/20/zuckerberg-apple-lesson-meta-100-billion-ai-bet/) Published by News Central Site: https://newscentral.site/investment-of-up-to-200-billion-in-ai-development-revealed/