内容来源：https://www.livemint.com/ai/artificial-intelligence/exgoogle-ceo-eric-schmidt-us-lagging-behind-china-ai-race-reasons-open-source-consumer-practical-applications-5-points-11759122133557.html

内容总结：

谷歌前首席执行官埃里克·施密特近日公开表示，在人工智能技术竞赛中，美国正落后于中国。他在9月25日的科技峰会上指出，虽然美国AI企业获得巨额融资和市场估值，但发展重心与中国存在显著差异。

施密特分析称，美国公司主要追逐通用人工智能的远景目标，而中国则更注重AI技术的实际应用转化。这种差异化发展使得中国AI技术通过开源模式更易推广，特别是在“一带一路”沿线国家获得广泛采用。他特别强调，中国企业在芯片禁令冲击下仍保持技术发展势头，并已形成自主技术体系。

值得关注的是，中国AI企业正将技术全面融入消费应用、机器人等日常领域。施密特以上海机器人企业为例，指出中国正将电动汽车领域的成功经验复制到机器人产业。他认为中国团队凭借务实作风和持续投入，正在这些领域形成竞争优势。

在技术路径方面，施密特指出中美选择存在根本差异：中国推崇开放权重和训练数据的开源模式，而美国主要采用封闭模式。这种差异可能导致全球多数国家最终选择中国AI模型。他特别提到深度求索公司的技术突破，称赞其在推理模型和强化学习方面取得重要进展，且以更低计算精度实现可比性能。

面对竞争态势，施密特建议美国在推进通用人工智能的同时，必须加强日常应用层面的技术竞争。他透露美国企业已开始调整策略，包括计划推出适配手机终端的轻量化开源模型，但强调美国仍需确保在移动端模型扩散中的主导地位。

（根据公开演讲内容整理）

中文翻译：

谷歌前首席执行官埃里克•施密特认为，尽管美国在人工智能领域投入巨额资金，但在这场科技竞赛中仍落后于中国。施密特在9月25日的"All-In"峰会上分享了他的观点，谈及中国如何在人工智能领域超越美国、美国公司需要将重点转向何处，以及西方可以采取哪些措施来追赶。

施密特指出，美国科技公司主要致力于实现通用人工智能，而中国则更专注于人工智能应用。他还指出，中国广泛使用开源项目和开源代码，这意味着他们的替代方案更容易获得，并可能被更广泛地采用。

施密特还承认，尽管中国人工智能领域的公司估值可能没有达到1000亿美元，但它们没有受到美国芯片禁令的严重打击，并且拥有与美国相媲美的自有技术。

当被问及美国如何在人工智能领域与中国竞争时，施密特回答说："我原本以为中国和美国在人工智能领域处于同一水平竞争，限制芯片的措施会拖慢他们的发展。（但）他们的做法与我想象的完全不同。"他指出，中国"没有追求疯狂的通用人工智能战略"，因为他们没有"深厚的资本市场"，无法"仅凭运气就筹集到1亿美元资金"。

他说："因此，他们非常专注于将人工智能应用到各个领域。我担心的是，在我们追求通用人工智能的同时，我们最好也能在日常应用方面与中国竞争。"施密特补充说，中国人工智能公司专注于"日常"事务，再加上他们"不可思议"的敬业精神，这种战略能够取得成功。

这位科技领袖指出，由于使用开源和开放权重模型，中国在大规模应用方面也处于领先地位，他称这是一个"重大的地缘政治问题"。他警告说："开源意味着开放代码。开放权重意味着开放训练数据。中国正在通过开放权重和开放训练数据进行竞争，而美国主要专注于封闭权重和封闭数据。这意味着世界上大多数国家，比如'一带一路'倡议参与国，将会使用中国模型而不是美国模型。"

施密特还承认，中国公司在技术方面取得了"重大成就"。他指出，尽管行业趋势是更加开放，但DeepSeek公司还是震撼了人工智能行业。施密特说："DeepSeek团队做得非常出色，不是吗？如果你看看DeepSeek的推理模型，特别是他们进行前后强化学习的能力，这是一项重大成就。而且他们似乎是用比美国模型更低的数值精度做到这一点的。"他指出，美国模型通常使用16位精度进行训练，而中国正在推进8位和4位精度。

当被问及美国公司是否需要采取措施与中国竞争时，施密特指出，相关计划已经开始实施。他说："一些大公司已经表示，他们希望成为开源领域的领导者。萨姆•奥尔特曼表示，O3模型的最小版本将会发布——我相信是开放权重的。他告诉我，这个模型更小，更容易训练，可以适配手机。"他总结说："所以一条路径是让超级计算机实现通用人工智能，但这将始终成本高昂等等。但我们也必须确保这些手持设备模型的扩散在美国的控制之下。无论是OpenAI、Meta、（谷歌的）Gemini还是其他公司。"

英文来源：

Former Google CEO Eric Schmidt feels that despite huge valuations and funding the United States is lagging behind China in the artificial intelligence (AI) tech race. Speaking at the All-In Summit on September 25, Schmidt shared his views on how China has out raced the US in AI, where American companies need to shift their focus, and what possible measures the west can take to catch up.
Schmidt noted that while US tech companies working in the AI field are largely pursuing artificial general intelligence or AGI, the Chinese are more focused on AI applications. He also noted that the overwhelming use of Open Source and Open Code projects from the Chinese means that their alternatives are more easily accessible and likely to be more widely adopted.
Schmidt also acknowledged that while Chinese companies in the AI space may not have $100 billion valuations, they have not been as hard hit by US imposing bans on chips and semiconductors, and have their own technologies matching American capabilities.
Responding to a question on how the US can compete with China on AI, Schmidt noted, “I had thought that China and the US were competing at the peer level in AI; and that the good work done to restrict chips, were slowing them down. (But), they're really doing something more different than I thought.”
He noted that China is “not pursuing crazy AGI strategies”, as they don not have “the depth of capital markets” and cannot “raise $100 million dollars on a wing and a prayer”.
“The result is, they're very focused on taking AI and applying it to everything. And the concern I have, is that while we're pursuing AGI, we better also be competing with the Chinese in day-to-day stuff,” he said.
Schmidt added that the Chinese AI companies are focused on “day-to-day” stuff and combined with their “incredible” work ethics, this strategy can win.
“They're very focused on taking AI and applying it to everything, (such as) consumer apps, robots and so on. I saw all the Shanghai robotics companies and these guys are attempting to do in robots what they've successfully done with electric vehicles,” he said.
Adding, “Their work ethics are incredible. They're well funded. It's not the crazy valuations that we have in America, but they can win across that.”
The tech leader noted that the Chinese also have in lead in wide scale adoption due to the use of open source and open weights models, calling it a “major geopolitical issue”.
“Open Source means open code. Open weights means to open training data. (And) China is competing with open weights and open training data and the US is largely and majority focused on closed weights and closed data. That means that the majority of the world, think of it, as the Belt and Road initiative are going to use Chinese models and not American models,” he warned.
Schmidt also acknowledged that Chinese companies have “major achievements” when it comes to the technical aspects. He noted that despite moves towards more open source, DeepSeek shook the AI industry.
“The DeepSeek people did such a good job, right? If you look at the reasoning model in DeepSeek and in particular their ability to do reinforcement, learning forward and back, this is a major achievement and it appears that they're doing it with less numeric precision than the American models,” Schmidt pointed out.
He noted that, while the American models are typically using 16-bit Precision for their training, the Chinese are pushing eight and four bit precision.
When asked if there were steps that American companies need to undertake in order to compete with China, Schmidt noted that the initiative has already begun, stating, “A number of the large companies have said that they want to be leaders in open source. Sam Altman indicated that the smallest version of the O3 model would be released — I believe, (it will be) open weights. He told me that this model is much smaller, much easier to train, and will or can fit on your phone.”
“So one path is to say that we'll have these super computers doing AGI, which will always be incredibly expensive and so forth. But we also have to watch to make sure that the proliferation of these models for handheld devices is under American control. Whether it's OpenAI or Meta, or (Google's) Gemini or what have you,” he ended.
Catch all the Business News, Market News, Breaking News Events and Latest News Updates on Live Mint. Download The Mint News App to get Daily Market Updates.