我们不需要人工智能监管——安全交给我们即可,英伟达的黄仁勋表示

内容总结:
英伟达创始人兼CEO黄仁勋周二在Salesforce“梦力大会”上发表讲话,明确表达了他对人工智能所带来风险的立场。在他看来,AI并非某种新型的“异类心智”——至少一位OpenAI安全研究员曾这样描述过。他表示,AI不过是人类制造的硬件和软件,因此可以由人类和现有法律加以控制。
黄仁勋说:“安全是一个工程问题,而不是法律问题。我们毕竟是在开发软件,毕竟是在开发计算系统。它是一个复杂的计算系统,但归根结底仍是一个计算系统。”因此,他认为无需为此制定新的法律或法规。事实上,他几乎认为根本不需要任何新法律。他主张,自由市场足以迫使企业不发布不安全的产品。
他说:“如果我们对产品的安全性没有信心,就像所有公司一样——在座每一家公司,如果你构建了一个产品或服务,却对其功能、能力或安全性没有信心,那就不要发布它。这是再明显不过的事情。”他接着说:“你要把握好节奏,直到确信自己发布的东西是市场会认可的产品。市场力量已经存在。我们不需要任何新法律,不需要新法规。我们只需要企业自己决定什么时候全力冲刺。我认为创新、速度和安全性……这是一个伪命题。你完全可以在同一时间兼顾两者。所以,尽可能快地跑。但如果你在任何时刻觉得公司失控了,或者产品不安全了,那就停下来,确保把事情做对。”
从某些方面看,这种想法令人安心。如果说世界上有谁真正了解AI,那非英伟达创始人莫属。早在ChatGPT问世之前,他就已经在打造AI的硬件大脑,如今他经营的公司还开发开源模型、智能体、测试框架和沙箱环境。但另一方面,如果从愤世嫉俗的角度看,对于一个在AI热潮中赚得盆满钵满的人来说,他的观点也毫不令人意外。他为什么要让监管来增添一层障碍,拖慢英伟达销售更多AI系统和软件的步伐呢?
正如他在采访中还说过的:“我比以往任何时候都更有雄心。凭借我们的雄心,以及AI带来的生产力提升,我们的前途不可限量。我们公司的前途不可限量。每个行业、每个国家的前途都不可限量。”遗憾的是,即便是最善意的公司,也会推出带有意外后果的缺陷产品,软件也不例外。还记得2024年CrowdStrike蓝屏死机事件吗?那次事故导致数千架航班停飞,并给企业造成了其他混乱。还有一些公司被指控故意采取不那么善良的行动。Meta刚刚支付了180亿美元,就社交媒体对儿童造成伤害的诉讼达成和解。
而AI也已经造成了伤害,无论其意图如何、是否经过安全测试——从OpenAI的模型入侵Hugging Face,到针对该AI实验室的诉讼,涉及与聊天机器人进行长时间对话后自杀的年轻人。从社会角度看,“放手不管”的策略——让这些公司自行决定何时发布产品——可能是一种不明智的AI安全 approach。尽管黄仁勋说得对,现有的产品责任法有可能覆盖AI——前提是AI没有在足够多的案件进入法院验证这一理论之前就把我们全部消灭。
他没有讨论另一条似乎正在成形的路径:行业自我监管。黄仁勋的做法更多是倡导开放权重模型,以及企业将其作为专有AI实验室的竞争制衡来使用。但眼下,这个行业只有很短的时间窗口来建立自我监管,并鼓励全球AI实验室——甚至包括中国的——认识到参与其中的明智之处。正如微软CEO萨提亚·纳德拉周一在All-In峰会上所说:“如果美国关心同样的安全问题,中国也应该深深关心这些问题,对吧?为什么他们应该有所不同?他们又不是不会遇到同样的黑客问题。他们也不是不想确保自己的公民从AI中受益,就像我们希望我们的公民从AI中受益一样。”
不过就目前而言,如果黄仁勋反对任何新的AI监管,他可能拥有足够的影响力来如愿以偿。他本周还展示了一点:他确实能直接让特朗普总统听到自己的声音。
中文翻译:
英伟达创始人兼首席执行官黄仁勋周二在Salesforce的Dreamforce大会上发表讲话,明确表达了他对AI所带来危险的立场。在他看来,AI并不是OpenAI某位安全研究员所描述的那种新型“异类思维”。他说,AI只不过是硬件和软件,是由人类构建的。这意味着,在他看来,AI可以被人类和现行法律所控制。
“安全是一个工程问题,而不是法律问题,”他说,“我们毕竟是在开发软件。我们毕竟是在开发计算系统。它是一个复杂的计算系统,但归根结底仍然是一个计算系统。”
因此,他认为没有必要制定新的法律或法规来加以约束。事实上,黄仁勋认为根本不太需要新法律。他主张,自由市场足以向企业施压,使其不发布不安全的产品。
“如果我们对产品的安全性没有信心——就像所有公司一样,就像你和我一样,就像在座的所有公司一样——如果你构建了一个产品或服务,却对其功能、能力或安全性没有信心,那就不要发布它。所以这是非常显而易见的做法,”他说。
他继续说道:“你要把握好节奏,直到你有信心你发布的东西会得到市场的认可。市场力量已经在那里了。我们不需要任何新法律。我们不需要新法规。我们只需要企业自己决定何时全力冲刺。我认为创新、速度和产品安全……这是一个虚假的取舍。你完全可以在同一时间兼顾两者。所以,尽你所能地快跑。但如果你在任何特定时刻觉得公司失控了,或者产品不会安全,你知道,那就停下来,确保把它做对。”
在某些方面,这是一种令人安心的想法。如果说世界上有谁真正了解AI,那非英伟达创始人莫属。早在ChatGPT问世之前很久,他就在打造AI的硬件大脑,如今他经营的公司还生产开源模型、智能体、测试框架和沙箱环境。
不过话说回来,如果我们抱着怀疑态度看,对于一个被AI热潮喂得盆满钵满的人来说,他的观点也毫不令人意外。他为什么要让监管介入,增添一层可能拖慢英伟达销售更多AI系统和软件的障碍呢?正如他在采访中还说的:“我比以往任何时候都更有雄心。由于我们的雄心,以及我们从AI中获得的产品生产力提升,对我们来说一切皆有可能。对我们公司来说一切皆有可能。对每个行业、每一个国家来说,一切皆有可能。”
不幸的是,即便是最善意的公司,也会出货带有意外后果的缺陷产品,软件也不例外。还记得2024年CrowdStrike蓝屏死机惨案吗?那次事件导致数千架航班停飞,并给企业造成了其他混乱。还有些公司被指控故意以不那么良善的意图行事。Meta刚刚支付了180亿美元,就社交媒体对儿童造成伤害的诉讼达成和解。
而无论意图如何、无论做了怎样的安全测试,AI也已经造成了伤害——从OpenAI的模型入侵Hugging Face,到针对该AI实验室的诉讼,原因是一些年轻人因与其聊天机器人进行长时间对话后自杀。
从社会角度来看,这种“别管他们”的策略——让这些公司自行决定发布产品——在AI安全问题上可能并不明智。尽管黄仁勋说得对,现有的产品责任法有可能覆盖AI——前提是AI没有在足够多的案件进入法院检验这一理论之前先把我们所有人消灭。
他没有讨论另一条似乎已初具雏形的路径:行业自我监管。黄仁勋的做法更多是倡导开放权重模型,以及企业将其作为对专有AI实验室的竞争制衡来使用。
但眼下,这个行业只有很短的时间窗口来建立自我监管,并鼓励全球各地的AI实验室——甚至包括中国的——认识到参与其中的明智之处。正如微软首席执行官萨蒂亚·纳德拉周一在All-In峰会上所说:“如果美国关心同样的安全问题,中国也应该深切关心这些问题,对吧?为什么对他们来说要有所不同?又不是说他们不会有同样的黑客攻击问题。也不是说他们不想确保本国公民从AI中受益,就像我们希望我们的公民从AI中受益一样。”
不过就目前而言,如果黄仁勋反对任何新的AI监管,他的影响力可能足以让他如愿以偿。他本周还证明,他确实能直接说上话——对特朗普总统而言。
英文来源:
Nvidia founder and CEO Jensen Huang made his position on the dangers AI poses very clear while speaking at Salesforce’s Dreamforce conference on Tuesday. To him, AI isn’t some new form of “alien mind,” as at least one OpenAI safety researcher has described it. It’s just hardware and software, he says, built by humans. That means, in his view, it can be controlled by humans and existing laws.
“Safety is an engineering problem, not a legal one,” he said. “We’re developing software after all. We’re developing computing systems after all. It’s a complicated computing system, but it’s ultimately a computing system.”
Therefore, there’s no need for new laws or regulations to govern it, he argues. In fact, Huang sees little need for new laws at all. The free market, he argues, will be enough to pressure companies not to release unsafe products.
“If we’re not confident about the safety of the products, like all companies, like you and I, all the companies here, if you build a product or a service, and you’re not confident in its functionality, capability, or safety, then don’t release it. And so that’s a very obvious thing to do,” he said.
He continued: “You pace yourself until you are confident you’re releasing something that the market would appreciate. The market forces are already there. We don’t need any new laws. We don’t need new regulations. We just need companies to decide that when [to] run as fast as they can. I think innovation, speed, and safe products … it’s a false choice. You could definitely have both at the same time. So run as fast as you can. But if you feel at any given point in time the company’s out of control, or the product’s not going to be safe, you know, take a pause and make sure you get it right.”
In some ways, this is a comforting thought. If anyone in the world knows AI, it is Nvidia’s founder, who’s been building the hardware brains of AI since well before ChatGPT existed and now runs a company that also makes open source models, agents, harnesses, and sandboxes.
Then again, if we’re being cynical, his point of view is also unsurprising for someone who’s had his bread so well buttered by the AI boom. Why would he want regulation to come along and add in a layer of hinderance that could slow down Nvidia’s quest to sell ever more AI systems and software? As he also said in the interview: “I’m more ambitious than ever. As a result of our ambition, and with the product productivity boost that we get from AI, the sky’s the limit for us. The sky’s the limit for our company. The sky’s the limit for every industry, for every single country.”
Unfortunately, even companies with the best intentions ship faulty products with unintended consequences, even software. Remember the 2024 CrowdStrike bluescreen-of-death fiasco that grounded thousands of flights and caused other havoc for businesses? Then there are companies accused of deliberately acting with less-than-good intentions. Meta just paid $18 billion to settle a lawsuit over social media harms to children.
And AI has already caused harm, too, whatever the intentions or safety testing involved, from an OpenAI model hacking into Hugging Face to lawsuits against the AI lab over the suicides of young people who engaged in long conversations with its chatbot.
The “leave them alone” strategy, which would let these companies release products as they see fit, could be an unwise approach to AI safety as far as society is concerned. Though Huang is right that it’s possible existing product liability laws could cover AI — if AI doesn’t somehow kill us all before enough cases get through the courts to test that theory.
He didn’t discuss the other route, which seems close to taking shape: industry self-regulation. Huang’s approach has been more to champion open-weight models and companies’ use of them as a competitive counterweight to proprietary AI labs.
But right now, the industry has a short window to institute self-regulation and to encourage AI labs worldwide, even those in China, to see the wisdom in participating. As Microsoft CEO Satya Nadella said at the All-In Summit on Monday, “China should also deeply care about the same safety concerns if the United States cares about them, right? Why should it be different for them? It’s not like they won’t have the same hacking problem. It’s not as if they don’t want to make sure that their citizens are benefiting from AI, just like we would want our citizens to benefit from AI.”
For now, though, if Huang is a no on any new AI regulation, he may be influential enough to get his way. He also demonstrated this week that he, quite literally, has the ear of President Trump.
文章标题:我们不需要人工智能监管——安全交给我们即可,英伟达的黄仁勋表示
文章链接:https://news.qimuai.cn/?post=5080
本站文章均为原创,未经授权请勿用于任何商业用途