从“永恒”到“不到一年差距”：谷歌前CEO施密特为何对中国AI的预判180度大转弯|中国|人工智能|埃里克|知名企业|美国|谷歌|赫尔穆特·施密特

谷歌前CEO埃里克-施密特（Eric Schmidt）11月25日接受采访时表示，中国在人工智能领域追赶美国的速度之快令他感到震惊，但是仅仅在半年前他可不说这么说的，我把施密特5月份的说法和11月的说法做一个对比，非常有意思

5月份

Eric Schmidt，

好消息是，美国远远领先于中国和其他国家，我认为这种情况在一段时间内都会持续下去。

对我来说，国家竞争力是未来10到20年的主要挑战，因为中国正专注于主导某些行业，我们需要与他们竞争，并确保我们能够胜出。

在人工智能领域，我们领先中国两到三年，可能更多，而在我的世界观中，这几乎是永恒的差距。我认为我们目前的形势相当不错。

我花了十年的时间试图说服欧洲真正去创新，而不是一味地监管，但他们仍然不断出台监管措施。当前的《人工智能法案》本质上是监管，而不是对未来的投资。

从中可以看出，欧洲很可能无法在这一领域发挥重要作用。中国当然也在因芯片短缺等问题而苦苦挣扎，尽管如果他们能获得所需的硬件，就可能取得胜利。

世界其他地方对此关注不够

英文原文：

The good news is the U.S. is way ahead of China and everybody else,and I think that's going to continue for a while.

To me,

National Competitiveness is the challenge for the next 10 or 20 years,because the Chinese are really focused on dominating certain industries,and we need to compete with them and make sure we win.

In the case of artificial intelligence,

we are well ahead two or three years, probably we have China,which in my world is in eternity.I think we're in pretty good shape.

I sent 10 years trying to convince Europe to actually innovate instead of regulate,and they just keep regulating.The current U.A.I.E.Act is essentially regulation, not investment in the future.You can see that Europe is highly unlikely to be relevant.

China, of course, is struggling because of chips shortages and so forth,although they're ready to win if they can get the harder that they need.

The rest of the world is not focused enough on this.

11月25日

Eric Schmidt，

本周，中国发布了两个开源库

一个是非常强大的问题求解器（应该指的是Deepseek最近发布的推理模型R1-Lite-Preview），另一个是一个大语言模型（应该指阿里巴巴的Qwen2.5-72B-instruct），在某些情况下，它的能力相当于甚至超越了 Meta 的模型，而这些模型是人们每天都在使用的，它被称为 Llama3-405B

当我看到这个消息时，我感到非常震惊，因为我之前在与中国的对话中曾假设，他们的进度会落后两到三年，但现在看来，他们可能只落后一年的时间

所以，公平地说，美国领先，而中国可能会在一年之内赶上

其他国家都远远落后了

现在，我并不是在暗示中国会对美国城市发起恶意攻击，但我确实认为有可能第三方会从中国，因为它是开源的，或者从美国窃取技术，如果他们掌握了这个水平的技术，就会这么做。

因此，威胁升级矩阵，随着每一次技术改进而上升

英文：

This week, there were two libraries from China that were released .Open source.One is a problem solver that's very powerful.And another one is a large language model that's equal to, in some cases,exceeds the one from meta with it's they use every day.It's called Llama3-405B

I was shocked when I read this,because I had assumed that in my conversation with the Chinese,that they were two to three years late,it looks to me like it's within a year now.

So be fair to say it's the US and then China within a year's time.

Everyone else is well-behive.

Now,I'm not suggesting that China will launch a rogue attack against us in American City.I am alleging that it's possible that a third party could steal from China,because it's open source, or from the US, if they're my level and do that.

So the threat, the threat escalation matrix,goes up with every improvement.

就我个人对施密特提到的两个模型的感觉和追踪来看，确实在国外引起了前所未有的重视，阿里的Qwen系列模型现在陆续被很多国外的团队采用并纳入默认支持模型，deepseek的R1表现可以看我这篇文章

著名计算机科学家，香港大学马毅教授更乐观，

不仅是开源领域，“我相信中国很快也会在智能基础研究方面遥遥领先，因为美国仍在崇拜蛮力式的方法”