Governments, companies, nonprofits should invest in free, open source AI [pdf]
291 points
• 1 day ago
• Article
Link
在 20 世纪 80 年代，作者曾与自由软件运动的创始人 Richard Stallman 就开放软件的利弊争论多年。起初作者为专有控制辩护，认为公司需要所有权才能推动技术进步，但他最终认识到，软件是一种重要的知识体系，只有通过透明才能不断壮大。这样的认识转变反映了更广泛的历史潮流：像 GNU/Linux 这样的开源项目证明，公开的集体协作能够胜过私营机构的封闭系统，并最终成为现代数字世界的承重基石。
如今，先进人工智能的兴起呈现出类似但规模更大的冲突。随着前沿 AI 模型越来越被企业壁垒所封锁，这一领域面临停滞的风险。 AI 正日益成为科学、专业和个人推理的基础设施，缺乏透明性令人担忧。当模型被少数公司当作黑箱掌控时，用户就无法审查其底层机制，只能依赖那些输出难以完全验证或理解的"神谕"。
反对开放 AI 的人常以安全和潜在滥用为由，担心公布模型方法等于散布危险工具。然而，作者认为真正的安全并非依靠模糊不清来实现。正如科学界在公开研究的前提下管理风险一样，AI 行业也应以透明为先。封闭系统仍然易受泄露和越狱的威胁，而权力集中于少数公司之手，会让这些企业单方面塑造未来的数字资源库，并为他人的知识获取制定规则。
当前所谓促进开放 AI 的努力常常流于表面——提供可运行模型的代码，却隐瞒构建模型所用的数据与方法。结果是用户只能得到未经解释就能运行的"魔数"，而非一个开放、可复现的基础设施。这种开放只是企业暂时施予的恩惠，而非全球社区可以独立研究或改进的可靠共享资源。
要走出困境，就必须重新承诺将开源 AI 作为公共产品来资助和支持。尽管私营部门的创新依然重要，但不应成为唯一选择。通过设立公共计算资源拨款、促进大学与非营利组织间的协作，并要求用公共资金开发的 AI 必须保持开放，社会可以阻止开放知识的流失。最初的开源运动已证明，共享的基础能够带来更快、更稳健的创新，我们必须在 AI 领域被永久关闭于公众之外之前，将这些经验教训付诸实施。
In the 1980s, the author spent years debating the merits of open software with Richard Stallman, the founder of the free software movement. While the author initially defended proprietary control, arguing that companies required ownership to advance technology, he eventually recognized that software represents a vital body of knowledge that grows stronger through transparency. This shift in perspective mirrored a broader historical trend, where open source initiatives like GNU/Linux proved that collective, public collaboration could outpace the closed systems of private entities, ultimately becoming the load-bearing foundation for the modern digital world.
Today, the emergence of advanced artificial intelligence presents a parallel conflict, though on a much larger scale. As frontier AI models are increasingly locked behind corporate walls, the field faces the risk of stagnation. Because AI is effectively becoming the primary infrastructure for scientific, professional, and personal reasoning, the lack of transparency is concerning. When models are treated as black boxes controlled by a few firms, users lose the ability to audit the underlying processes, forcing them to rely on "oracles" whose outputs cannot be fully verified or understood.
Critics of open AI often cite safety and the potential for misuse, fearing that publishing the methods behind these models is equivalent to distributing dangerous tools. However, the author argues that true security is not achieved through obscurity. Just as the scientific community manages risks while keeping research public, the AI industry should prioritize transparency. Closed systems remain vulnerable to leaks and jailbreaks, while the concentration of power in a few companies allows those entities to unilaterally shape the digital library of the future, dictating the terms of knowledge access for everyone else.
Current efforts to promote open AI often fall short, providing the code to run a model while concealing the data and methods used to build it. This creates a state where users are left with "magic numbers" that function without explanation, rather than an open, replicable foundation. This form of openness is merely a temporary favor granted by corporations, rather than a reliable, shared resource that can be independently studied or improved by the global community.
The way forward requires a renewed commitment to funding and supporting open source AI as a public good. While private sector innovation remains important, it should not be the sole option available. By establishing public compute grants, fostering collaboration between universities and nonprofits, and mandating that AI developed with public funds remains open, society can prevent the erosion of open knowledge. The success of the initial open source movement proves that shared foundations lead to faster, more robust innovation, and it is imperative that we apply these lessons before the field of AI becomes permanently closed to the public.
106 comments • Comments Link
• 面向 open-source AI models 的定向激励计划，如果以严格的 VRAM 和硬件限制为约束来设计，可能有助于培育创新并让规模较小的参与者获得关注。
• 要有效实施此类竞赛，需要封闭且定期更新的基准测试以防作弊，但创建和维护高质量评估既昂贵又复杂。
• open-source AI 很难复制像 Linux 那样的软件贡献模式，因为训练 frontier models 需要大规模、集中的资本投入，这超出兼职贡献者或个人的能力范围。
• 虽然 commodity software 往往在开源生态中蓬勃发展，但那些尚未被发表的前沿研究或需要庞大基础设施的领域，仍由有资源的商业实体主导。
• Frontier AI 的开发集中在少数由亿万富翁支持的组织手中，这带来了形成强大寡头的风险，因此 open-weights models 对于维持更广泛、更民主的知识分配至关重要。
• 市场力量可能最终会把 AI tokens 商品化，类似于 mobile data，那样大型实验室可能会更多地专注于微调和部署，而非囤积核心智能。
• 对政府出资奖励池存在质疑，一些人认为这些资金更应投入到基础公共需求，而不是补贴竞争性的业余爱好。
• 当前的 U.S. healthcare system 背负着极高的行政成本、过度检查以及"客户永远正确"的文化，导致支出高昂却常常得不到最佳结果。
• 支持 nationalized healthcare 的论点，经常与对政府官僚主义、资源配给风险的担忧产生冲突，人们担心系统性变革不仅仅是消除企业利润就能解决的复杂权衡。
• 关于公共支出的争论反映出两种张力：一方面人们渴望高效的国家主导服务，另一方面又担心政府管理会重演诸如 schools 或 social security systems 等现有公共机构中的低效。
这场讨论体现了人们对 AI 未来以及 American social contract 可持续性的深层焦虑。一方面，关于通过奖金和硬件分级竞赛来培育 open-source AI ecosystem 的技术性论述，仍受限于训练 frontier models 所需的昂贵、难以去中心化的现实；另一方面，谈话转向更广泛的社会政治不满，尤其集中在医疗保健和财富分配问题上。在这里，对当前以高成本、"overcare"以及企业中间商权力为特征的系统的挫败感，与对政府是否能成为更有效或更具同情心管理者的深刻怀疑发生碰撞。归根结底，参与者们在探讨公共投资是否能解决这些复杂的技术与社会问题，还是这些系统注定低效并会被特定利益集团绑架。 • A targeted inducement prize program for open-source AI models, structured around strict VRAM and hardware constraints, could foster innovation and help smaller players gain visibility.
• Implementing such competitions effectively requires closed, periodically refreshed benchmarks to prevent gaming, though creating and maintaining high-quality evaluations is an expensive and complex undertaking.
• Open-source AI struggles to replicate the contribution model of software like Linux because training frontier models requires massive, centralized capital expenditure that goes beyond what part-time contributors or individuals can provide.
• While commodity software often thrives under open-source models, non-commodity domains—where state-of-the-art research is unpublished or requires massive infrastructure—remain dominated by commercial entities with the resources to sustain development.
• The concentration of frontier AI development within a few billionaire-led organizations creates risks of a powerful oligarchy, making open-weights models essential for maintaining a broader, more democratic distribution of knowledge.
• Market forces may eventually drive the commoditization of AI tokens, much like mobile data, potentially leading to a landscape where large labs focus on fine-tuning and deployment rather than exclusively hoarding core intelligence.
• Skepticism exists regarding government-funded prize pools, with some arguing that such capital should be directed toward fundamental public needs rather than subsidizing competitive hobbyism.
• The current U.S. healthcare system suffers from extreme administrative overhead, over-testing, and a "customer is always right" culture, leading to outcomes that are expensive and often suboptimal despite high spending.
• Arguments for nationalized healthcare often clash with concerns about government bureaucracy, the potential for resource rationing, and the reality that systemic change requires difficult trade-offs that go beyond merely eliminating corporate profits.
• Debates over public spending reflect a tension between the desire for efficient, state-led services and the fear that government administration will replicate the inefficiencies seen in existing public institutions like schools or social security systems.
The discussion reflects a deep anxiety regarding the future of AI and the sustainability of the American social contract. On one hand, there is a technical discourse about how to foster an open-source AI ecosystem through prizes and hardware-tier competitions, though this is tempered by the reality that training frontier models is inherently expensive and difficult to decentralized. On the other hand, the conversation shifts toward broader sociopolitical grievances, particularly regarding healthcare and wealth distribution. Here, the frustration with current systems—characterized by excessive costs, "overcare," and the power of corporate middlemen—is met with a deep-seated cynicism regarding the government's ability to act as a more efficient or benevolent steward. Ultimately, the participants are grappling with whether public investment can solve complex technical and social problems, or if those systems are inherently destined to be inefficient and captured by specific interest groups.