TThe technology industry is currently embroiled in a heated debate over the future of AI: should powerful systems be open source and freely accessible, or closed and rigorously monitored for dangers?
On Tuesday, Meta CEO Mark Zuckerberg fired a salvo in this ongoing battle, publishing not only a new series of powerful AI models, but also a manifesto vigorously defending the open source approach. The document, widely praised by venture capitalists and technology leaders like Elon Musk It is Jack Dorsey, serves as both a philosophical treatise and a rallying cry for proponents of open source AI development. It comes at a time when intensifying global efforts to regulate AI have galvanized resistance from open source advocates, who see some of these potential laws as threats to innovation and accessibility.
In the heart of Meta announcement on Tuesday was the launch of its latest generation of Llama large language models, the company’s answer to ChatGPT. The biggest of these new models, says Meta, is the first major open-source language model to reach the so-called “frontier” of AI capabilities.
Meta has taken a very different strategy with AI compared to its competitors OpenAI, Google DeepMind and Anthropic. These companies sell access to their AI through web browsers or interfaces known as APIs, a strategy that allows them to protect their intellectual property, monitor the use of their models and prevent bad actors from using them. On the other hand, Meta has chosen to open source the “weights”, or underlying neural networks, of its Llama models – meaning they can be downloaded for free by anyone and run on their own machines. This strategy put Meta’s competitors under financial pressure and won many fans in the software world. But Meta’s strategy has also been criticized by many in the AI safety field, who warn that open sourcing powerful AI models has already led to social harms like deepfakes, and could in the future open a Pandora’s box of worse dangers.
In his manifest, Zuckerberg argues that most of these concerns are unfounded and frames Meta’s strategy as a democratizing force in the development of AI. “Open source will ensure that more people around the world have access to the benefits and opportunities of AI, that power is not concentrated in the hands of a small number of companies, and that the technology can be implemented more uniformly and securely across all the countries. society,” he writes. “It will make the world more prosperous and safer.”
But while Zuckerberg’s letter presents Meta as being on the side of progress, it’s also a deft political move. Recent polls suggest that the American public would welcome laws that restrict the development of potentially dangerous AI, even if it means hindering some innovation. And several pieces of AI legislation around the world, including the SB1047 bill in California and the ENFORCE Act in Washington, D.C., would impose limits on the types of systems that companies like Meta can open, due to security concerns. Many of the venture capitalists and technology CEOs who celebrated Zuckerberg’s letter after its publication have in recent weeks mounted a growing campaign to shape public opinion against regulations that would restrict the release of open source AI. “This letter is part of a broader trend of some Silicon Valley CEOs and venture capitalists refusing to take responsibility for the harm their AI technology may cause,” says Andrea Miotti, executive director of the security group AI Control AI.
The philosophical underpinnings for Zuckerberg’s commitment to open source, he writes, stem from his company’s long fight against Apple, which through its iPhone operating system restricts what Meta can build, and which through its App Store becomes with a share of Meta’s revenue. He argues that building an open ecosystem – in which Meta’s templates become the industry standard due to their customizability and lack of restrictions – will benefit both Meta and those who rely on its templates, while only harming companies. that seek to monetize and that intend to retain users. (Critics to pointhowever, that Llama models, while more affordable than their competitors, still come with usage restrictions that fall short of true open source principles.) Zuckerberg also argues that closed AI vendors have a business model that relies from selling access to its systems — and suggests that its concerns about the dangers of open source, including governments’ lobbying against it, may result from this conflict of interest.
Responding to security concerns, Zuckerberg writes that open source AI will be better at addressing “unintentional” types of harm than the closed alternative, due to the nature of transparent systems being more open to scrutiny and improvement. “Historically, open source software has been more secure for this reason,” he writes. As for intentional harm, such as misuse by bad actors, Zuckerberg argues that “large-scale actors” with high computational resources, such as companies and governments, will be able to use their own AI to police “less sophisticated actors” who misuse of open source systems. . “As long as everyone has access to similar generations of models – promoted by open source – governments and institutions with more computing resources will be able to check bad actors with less computing,” he writes.
But “not all ‘big players’ are benevolent,” says Hamza Tariq Chaudhry, a US policy expert at the Future of Life Institute, a nonprofit focused on AI risk. “More authoritarian states are likely to repurpose models like Llama to perpetuate their power and commit injustice.” Chaudhry, who is originally from Pakistan, adds: “Coming from the Global South, I am acutely aware that AI-fueled cyberattacks, disinformation campaigns and other harms pose a much greater danger to countries with nascent institutions and severe resource constraints, far from Silicon Valley.”
Zuckerberg’s argument also doesn’t address a central concern of many people concerned about AI security: the risk that AI will create an “attack-defense asymmetry,” or, in other words, strengthen attackers while doing little to strengthen defenders. “Zuckerberg’s statements show a worrying disregard for basic security in Meta’s approach to AI,” says Miotti, director of Control AI. “When dealing with catastrophic dangers, it is a simple fact that the offense only needs to get lucky once, but the defense needs to get lucky every time. A virus can spread and kill in days, while implementing a treatment can take years.”
Later in his letter, Zuckerberg addresses other concerns that open source AI will allow China to gain access to the most powerful AI models, potentially harming U.S. national security interests. He says he believes that closure models “will not work and will only harm the US and its allies”. China is good at espionage, he argues, adding that “most technology companies are nowhere near” the level of security that would prevent China from stealing heavyweights from advanced AI models. “It seems more likely that a world of only closed models will result in a small number of large companies, in addition to our geopolitical adversaries, having access to leading models, while startups, universities and small companies miss out on opportunities,” he writes. “Furthermore, restricting American innovation to closed development increases the likelihood that we will not lead.”
Miotti is not impressed by the argument. “Zuckerberg admits that advanced AI technology is easily stolen by hostile actors,” he says, “but his solution is to simply provide it for free.”
This story originally appeared on Time.com read the full story
