Meta's AI Strategy: Open-Sourcing LLaMA 3.1 and Its Impact on the AI Landscape

Meta AI Strategy - Give Something To Get Something

Zuckerberg's firm is spending a lot more than Alphabet did: $120 billion in capital expenditure over the next three years. Are the biggest technology firms overinvesting in artificial intelligence? Probably. Meta Platforms' (META META:CA) $40 billion in capital spending this year, for example, will be around double what it allocated in 2021. Shareholders have been broadly supportive of that. But what happens if companies give away the results of their investment for free? That's effectively what CEO Mark Zuckerburg is promising to do. His $1.1 trillion company says that its new AI models - the latest, Llama 3.1, was released on Tuesday - will be given to developers to tweak and freely use, plugging in their own data without having to share it with Meta. That's a divergence from rivals who are pursuing more proprietary strategies. Microsoft-backed OpenAI, for example, is charging for access to its own technologies. This might sound like a giant involuntary giveaway on the part of the shareholders who effectively fund Meta's AI investment. The theory, though, is that Meta will end up long-term richer for it. Zuckerberg says that putting Llama into as many hands as possible will speed up the development of new products and tools, and lower associated costs. There's something to this argument. Having multiple people fooling around with AI will no doubt generate unexpected uses. It should also strike a competitive blow against rivals like OpenAI whose products are effectively paywalled. One encouraging precedent is Android, the smartphone operating system owned by Alphabet. The parent company of Google bought the bones of Android for an estimated $50 million in 2005, a deal a Google executive later called the company's best ever. Google (GOOG, GOOGL) licensed the software for free, but initially insisted phones have the company's online store, search and browser preinstalled. That eventually generated regulatory pushback, but cemented the company's dominance in mobile search and underpins its $2.1 trillion valuation. Zuckerberg's firm is spending a lot more than Alphabet did: $120 billion in capital expenditure over the next three years, according to analyst estimates from LSEG. For that to generate a return of, say, 20%, Meta will have to add $24 billion of annual earnings. Such a sum would be equivalent, based on its current profit margin, to nearly $75 billion of revenue - roughly half again on top of what it is expected to make this year. The question is where so much extra revenue will come from, if not from charging for AI models. That's something even the souped-up Llama might struggle to answer. Meta Platforms released its Llama 3.1 artificial intelligence models on July 23, and said that it would follow an "open source" strategy where developers can use and modify its models free of charge. CEO Mark Zuckerberg said that Llama 3.1, which can translate and answer questions in eight languages, answer more complex queries and write computing programs, is competitive with most advanced AI offerings produced by rivals. He added that he expects Meta's future models to become the most advanced in the industry.

How Llama 3.1 is Saving Zuckerberg's Metaverse

"Metaverse was always going to be a very long-term thing. Some things have been better than I thought; some have gone slower," said Mark Zuckerberg. Meta continues to dominate the open-source market with its latest powerful open-source model Llama 3.1 405B. The developer circle has already gone berserk over the model's enhanced capabilities, which is said to surpass OpenAI's GPT-4o. The new model is another promising step in accelerating CEO Mark Zuckerberg's metaverse dream. Zuckerberg's vision of making an AI social media falls in line with the current creator market. "There are almost 200 million creators on our platforms. They are all trying to build their community. People want to interact with them, and there aren't enough hours in the day," said Zuckerberg in a recent interview with Bloomberg. Zuckerberg believes that an AI version will solve this problem. "I want to make it so that every single one of them can easily train an AI version of themselves; they can make it what they want," he said. He also believes that through an AI version, the artist or creator will have control over how they would interact with their community. Meta announced that the latest model Llama 3.1 405B is integrated into Meta AI on Whatsapp and Meta.ai. Furthermore, the Meta AI chatbot will now support Hindi too. Source: LinkedIn With this vision, Zuckerberg is promoting the very concept on which any form of social media works -- engagement. As of 2024, Instagram users spend an average of 33 minutes per day on the platform, which was 15 minutes per day in 2019. However, the concern of the platform losing users has also come up. The new AI mode could be the next best way to keep Meta's users hooked to the platform. "I think that's going to be great and there's going to be millions or, eventually hundreds of millions of those," he said. With Meta rolling out a number of product features in the past many months, the likes of AI stickers and digital avatars, the company has only found more ways to keep the interaction going. "I think that people want to connect with each other. I actually think that all the other stuff is generally noise. But, technology allows you to do that in a better way," said Zuckerberg. "The metaverse thing was always going to be a very long-term thing. I think some things have been better than I thought. Some have gone slower," said Zuckerberg when asked about his vision and performance of metaverse. Meta has been building a well-connected web of metaverse via their products over the last few years. With AI in the picture, Zuckerberg believes it has surpassed his assumptions. "If you asked me five years ago, I would've guessed that AR would come before AI, not the other way around. So, some of this is just kind of setting yourselves up to ride the different waves when they come in," he said. Zuckerberg also remains positive about the trend in which tech has evolved. Referring to Facebook, which was text-heavy when released, he said it brought in photos and videos as smartphones and improved mobile networks came into the picture. "I just don't think that's the end of the line. I think it gets more immersive." The latest Ray Ban Meta glasses are a perfect example of metaverse working well, especially after the AI acceleration. The first version of the smart glasses, which was released three years ago, missed the mark. However, with Meta AI's integration with the eyewear that allows one to have an immersive and seamless interaction, the glasses were back in vogue. "They're stylish. They're good glasses. It's a great form factor for AI," said Zuckerberg. While the possibility of metaverse's resurgence is evident, Zuckerberg is also aware that they will not be the only ones to come up with it, a philosophy that supports Meta's open-source motive. "A pretty core part of our philosophy is we don't believe there's going to be, like, one AI to rule them all. Our vision is that there's going to be millions or just billions of different models out there," said Zuckerberg. In the context of organisations attempting to create a super-intelligent entity, Zuckerberg compares them to playing God. This perspective is viewed as both incorrect and unappealing.

The AI race's biggest shift yet

The AI race is quickly changing. Focus is shifting away from the models themselves to the products they power, as evidenced by the events of this week. First, there was Mark Zuckerberg continuing his scorched-earth campaign to drive down the cost of accessing foundational AI models. Llama 3.1 narrows the gap between the performance of open- and closed-source AI, and Meta claims it costs roughly half as much as OpenAI's GPT-4o to run. In a video announcing the news, Zuckerberg wore a custom shirt with a quote from Emperor Augustus emblazoned in Latin: "At the age of nineteen, on my own initiative and at my own expense, I raised an army."

Interview with Mark Zuckerberg about AI open source and future platforms

Mark Zuckerberg discusses his vision for the future of AI and Meta's strategic direction. He emphasizes the importance of open-source AI, the evolution of Meta's technology, and the broader implications of AI on society and business. Zuckerberg also touches on personal aspects of his life, his leadership philosophy, and his views on the role of technology in the future. The CEO of Meta, envisions a future where artificial intelligence (AI) plays a pivotal role in both society and business. His focus on open-source AI, technological evolution, and the broader implications of AI underscores Meta's strategic direction. Zuckerberg's insights and Meta's AI initiatives have the potential to shape the future of digital experiences and drive innovation across various domains. Meta's AI strategy hinges on its commitment to open-source AI. Unlike closed-source competitors like OpenAI and Google, Meta champions open-source initiatives to drive innovation and collaboration. This approach fosters transparency and community involvement, creating a more inclusive technological ecosystem. By embracing open-source AI, Meta aims to: Meta's dedication to open source dates back to 2003, with projects like Open Compute highlighting its commitment to sharing technological advancements with the broader community. This historical context is crucial for understanding Meta's current AI initiatives and its vision for the future. Zuckerberg's transformation, often dubbed the "Zuckaissance," reflects his ability to adapt to the evolving digital landscape. By embracing public perception and memes, he has reshaped his leadership style to stay relevant and influential in the tech industry. This personal and professional reinvention demonstrates Zuckerberg's commitment to staying at the forefront of technological advancements and driving Meta's strategic direction. Here are a selection of other articles from our extensive library of content you may find of interest on the subject of Llama 3 : Meta aims to gain control over future technology platforms by focusing on AI and augmented/virtual reality (AR/VR). This vision includes developing AI models that enhance user interactions and offer innovative solutions for businesses and creators. By investing in innovative AI research and development, Meta seeks to: Meta AI has seen significant advancements, with applications across various domains. The goal is for Meta AI to become the most used AI assistant, providing personalized and efficient support to users. This aligns with Meta's broader vision of integrating AI into everyday life, making technology more accessible and user-friendly. The introduction of Llama 3.1, a model with 405 billion parameters, marks a milestone in Meta's AI development. These large models train smaller, specialized models, allowing more precise and efficient AI applications. This approach underscores Meta's belief in the value of multiple AI models tailored to specific tasks, rather than relying on a single dominant AI. Zuckerberg sees AI as a versatile tool for creators and businesses, emphasizing the adaptability of AI technologies and their potential to drive innovation across various industries. By providing tools that enhance creativity and productivity, Meta aims to empower users and businesses to: The development of the Metaverse is a key component of Meta's long-term vision. Integrating AI with AR and VR technologies presents both opportunities and challenges. Progress in this area will shape the future of digital experiences, offering immersive and interactive environments for users. To realize this vision, Meta is investing in: Meta's investment in AI infrastructure is crucial for maintaining a competitive edge in the rapidly evolving tech industry. However, there is a risk of overbuilding, which could lead to inefficiencies. Balancing investment with strategic planning is essential for sustaining growth and innovation in the AI sector. Meta's approach to AI investment includes: The pursuit of Artificial General Intelligence (AGI) represents a significant ambition for Meta. AGI involves creating AI systems capable of performing a wide range of tasks, matching or surpassing human intelligence. Achieving AGI would mark a transformative milestone in AI development, with far-reaching implications for society and technology. While the path to AGI is complex and uncertain, Meta's commitment to open-source AI and collaborative research brings us closer to this goal. Zuckerberg remains optimistic about the future of technology, particularly AI's potential to enhance creativity and productivity. His long-term vision includes a commitment to technological evolution and its positive societal impact. By driving innovation in AI, AR/VR, and other emerging technologies, Meta aims to create a future where technology empowers individuals, connects communities, and solves complex challenges. Mark Zuckerberg's vision for AI and Meta's strategic direction encompasses a broad range of initiatives and goals, from open-source AI development to the integration of AI with AR/VR. By focusing on innovation, transparency, and collaboration, Zuckerberg and Meta are poised to shape the future of digital experiences and drive transformative change across industries. As AI continues to evolve and permeate various aspects of our lives, Meta's commitment to responsible development and its pursuit of AGI will play a crucial role in shaping the trajectory of this powerful technology.

Why is Mark Zuckerberg giving away Meta's crown jewels?

As alter egos go, Augustus Caesar is not a bad one for Mark Zuckerberg, pontifex maximus of Meta, owner of the Facebook family of apps. Both men started their march to power as teenagers. Both stopped at nothing to build empires -- though unlike the impetuous Mr Zuckerberg, Augustus's motto was "make haste slowly". Both gave the illusion of sharing power (Augustus with the Senate, Mr Zuckerberg with shareholders) while wielding it almost absolutely. The Roman emperor is Mr Zuckerberg's role model. In a recent podcast he used the 200-year era of stability ushered in by Augustus to illustrate why he is making Meta's generative artificial-intelligence (AI) models available in a way that, with some poetic licence, he calls open source. On July 23rd Mr Zuckerberg issued a manifesto laying out in greater detail the business case for open-source AI. That coincided with the release by Meta of Llama 3.1, a freely available large language model (LLM) whose most powerful version, it says, rivals the top offering from OpenAI, maker of ChatGPT. Mr Zuckerberg said Meta's intent was to liberate itself from the sort of gatekeepers that have constrained it in the past, such as Apple and its iPhones. That sounds sensible. It was lost on no one, though, that Meta is Llama's sole gatekeeper. Meta's new model is certainly an attention-grabber. The biggest version has 405bn parameters (a common definition of LLM power), almost six times those in its predecessor. Mr Zuckerberg claimed that by next year the company's models will reign supreme, throwing down the gauntlet to rivals like OpenAI that have taken a closed approach. As both open and closed models get bigger, the debate over which is better is developing an almost theological intensity. On one side are the open-source purists in favour of decentralised "little tech". On the other are closed-source realists who argue that greater centralisation and control are better for safety and national security. Mr Zuckerberg's manifesto further stirs that debate. Though questions remain about how genuinely open Meta's models are, and its commitment to the approach, he makes a good case. As he points out, open-source software has an illustrious pedigree. In the 1990s Linux, an obscure operating system created by a university student, eventually became the industry standard for servers, thanks in part to the backing of IBM, a tech giant of its day. The beauty of Linux's approach was that it provided full access to its source code, enabling developers to modify and improve it. That differs subtly from Meta's approach to AI. Percy Liang, co-founder of Together AI, a cloud-computing startup that will use Llama 3.1, calls the tech giant's models "open weights", rather than open source. Meta makes available the numerical values used in its models, known as weights, but doesn't reveal the data on which the models are trained, which is the equivalent of the source code. That may reduce the ability of developers to customise its models. It is better than nothing, though. This also raises the question of whether or not Meta might change its approach, leaving developers that rely on its models high and dry. Meta is not a charity, and building LLMs can be costly. Investors have shown in their hostile reaction to Mr Zuckerberg's metaverse ambitions that, despite his control of the company's voting shares, he does not have a blank cheque to splurge on whatever he likes. If Meta does not get the commercial benefits it expects, it may be forced to reconsider its approach. Openness, meanwhile, raises two big safety concerns. The first is harm-prevention. Though Meta has probed Llama 3.1 for dangers, the bigger models get, the more risk there is that they could go rogue or be misused. Once released, such models do not have a kill switch. That, in turn, raises the issue of liability. Who bears responsibility if these models fall into the hands of bad actors? Regulators are grappling with such questions; a clampdown could affect the long-term viability of open-source AI. Mr Zuckerberg's rebuttal starts with self-interest. Meta benefits from its interactions with the open-source community, which will suggest ways to make its models better, he argues. Better models should in turn help the firm improve the performance of the AI products it offers to users of Facebook, Instagram and WhatsApp, boosting engagement and profit. Meta's business is based on advertising, rather than subscriptions, so it has no risk of cannibalising itself. What is more, though he does not say this, making its large language models available for free helps commodify the industry, undercutting the prospects of rival tech giants. As with IBM, which backed Linux against Microsoft's Windows, Meta's megabucks and clout are giving open-source AI a tailwind. Big firms such as Nvidia, creator of generative-AI chips and provider of related services, and Amazon Web Services, a cloud provider, are incorporating Llama 3.1 into their products. Mr Zuckerberg also insists that it is safer to have power concentrated in the hands of the many rather than the few. When it comes to national security, closing models to prevent China getting its hands on them would be counter-productive, he writes. It would hurt American innovation, and China might be able to steal the secrets anyway. Mr Zuckerberg's long-term bet is that openness will be good for the world as well as Meta. He has likened it to Augustus's Pax Romana. After years of civil war in the wake of Julius Caesar's death in 44BC, few in Ancient Rome could conceive of the idea of a prolonged period of peace and prosperity. Likewise, he reckons, few investors at this point are able to imagine the long-term potential of Meta giving away its crown jewels. Like his hero, he has battles to fight before his position is secure. But he is on the warpath.

Sam Altman, under fire from Elon Musk, has now offered his own vision of open-source AI

This story is available exclusively to Business Insider subscribers. Become an Insider and start reading now. Have an account? Log in. But what's notable in Altman's proposal was how he highlighted the role open-source AI models could have in strengthening America's hold on the field. "Making sure open-sourced models are readily available to developers in those nations will further bolster our advantage," Altman wrote. "The challenge of who will lead on AI is not just about exporting technology, it's about exporting the values that the technology upholds." Musk has repeatedly criticized Altman's leadership since he left the ChatGPT maker's board in 2018. Musk has been pummeling Altman based on his stance that OpenAI's systems aren't as open as they should, ideally, be. "OpenAI was created as an open source (which is why I named it 'Open' AI), non-profit company to serve as a counterweight to Google, but now it has become a closed source, maximum-profit company effectively controlled by Microsoft," Musk wrote in an X post in February 2023. "Not what I intended at all," he added, Musk filed a lawsuit against OpenAI in February this year, where he accused the company of violating its nonprofit mission by partnering with Microsoft. He later withdrew his lawsuit in June. To be sure, Musk also has some reservations about how open AI models should be. In private correspondence, the Tesla and xAI chief appeared to agree with OpenAI's then-chief scientist, Ilya Sutskever, that there should be limits to openness. "As we get closer to building AI, it will make sense to start being less open," Sutskever said in an email to Musk in January 2016. "The Open in OpenAI means that everyone should benefit from the fruits of AI after its built, but it's totally OK to not share the science (even though sharing everything is definitely the right strategy in the short and possibly medium term for recruitment purposes)," Sutskever continued. "Yup," Musk replied. Representatives for Altman and Musk didn't immediately respond to requests for comment from Business Insider sent outside regular business hours. Altman's position on open-source AI might also be less of a mea culpa to Musk, as it is a recognition that it could be where the industry is headed. On Tuesday, Meta CEO Mark Zuckerberg unveiled his company's latest AI model, Llama 3.1. The model, which Meta says can outperform OpenAI's GPT-4o, is open source and has been made available for the public to use for free. "I believe that open source is necessary for a positive AI future," Zuckerberg wrote in a blog post on the same day. "There is an ongoing debate about the safety of open source AI models, and my view is that open source AI will be safer than the alternatives. I think governments will conclude it's in their interest to support open source because it will make the world more prosperous and safer," Zuckerberg wrote.