世界经济论坛举办“人工智能入门”研讨会

寒风凛冽的周日夜晚，黄昏悄悄降临瑞士达沃斯。大家在一年一度的世界经济论坛的前夕静静地却有所保留地期待着将会发生的一些引人注目的事情——或者至少据说会发生的事情，即世界经济论坛将就我们如何在全球应对人工智能发表高论。

## We have to agree somewhere

我们必须在某个点达成一致

Certainly, we are nowhere near an agreed-upon position, nor do we even seem to be near a common way of looking at AI. But a start is better than nothing at all, and the World Economic Forum is perhaps the most appropriate place for it. Of all the voices and forces in and around AI, this congress is the one body that, in the opinion of this writer, has the most to say (at least for now) with nothing to gain: not materially or any other way.

当然，我们离达成一致立场还差得远，我们甚至连怎么看待人工智能的都未形成共识。但开个头总比什么都没有要好一些，世界经济论坛也许是最合适开头的地方。在笔者看来，世界经济论坛是所有人工智能领域的声音和力量中最有发言权的一个机构（至少目前如此），而且世界经济论坛本身并不能从中获益：无论是物质上还是其他方面。

It’s not as if AI is being discussed for the first time, but the gigantic leaps in the field in the last year alone – not to mention the much greater, almost unimaginable, expectations for this year and beyond – bring AI to a commanding position – permanently.

这一次大致也不是第一次讨论人工智能，不过仅去年一年内，人工智能领域取得了巨大飞跃，今年及以后的更大飞跃就更不用说了，肯定是几乎难以想象的飞跃期望——所有这些飞跃将令人工智能占据了主导地位，永久性地。

Why permanently? As I’ve written on multiple occasions in this venue, artificial intelligence will be the undoubtedly biggest advancement ever made by humans, given its potential to wrest superiority from us, frightful and fanciful is it now may seem. But it’s been discussed, so it shouldn’t be a shock. If any of us hasn’t shuddered at the scene in _2001: A Space Odyssey_ – where HAL9000 said, “I’m sorry, Dave, I’m afraid I can’t do that” – then my point about superiority is moot.

为什么是永久性呢？正如笔者多次撰文所言，人工智能无疑将是人类有史以来取得的最大进步，因为人工智能有可能会超越人类，尽管现在看来这似乎有些可怕和虚幻。但已经有过不少这样的讨论，所以不应该感到震惊。我们看到电影《2001：太空漫游》（2001: A Space Odyssey）里听到 HAL9000 说“戴夫，对不起，我恐怕不能那样做”那一幕时，如果我们没有感到不寒而栗的话，那么我说的人工智能超越人类的观点也就毫无意义了。

## **The Two Conflicts Over AI**

关于人工智能的两大冲突

As I see it, two major points of contention have emerged and taken front-and-center positions.

笔者觉得，主要争论点有两个并占据了前排中心位置。

## **1\. Technology advances vs. ethics**

1、技术进步与伦理

Always an issue with every technology advance ever made, from stone tools to AI, we humans have never failed to figure out not only the beneficial uses, but also the destructive ways. The reason, stated simplistically, is that we are more committed to what we could do than what we should do. Case in point: the New York Times’ copyright infringement lawsuit against Microsoft and Open AI for “scraping” intellectual property of the New York Times to use in teaching AI machines to learn. Neither a lawyer nor a publisher, I’ll not offer an opinion, but as an observer and writer, I submit there’s something wrong in there somewhere – and it appears awfully big.

从石器到人工智能，每一次技术进步都会遇到一个问题，我们人类往往没有忘记不仅要找有益的用途，也要找毁灭的方式。简单地说，原因在于我们更关注我们能做什么，而不是我们应该做什么。来看这个案例：《纽约时报》告微软和OPENAI公司版权侵权，指控两家公司“爬”《纽约时报》的知识产权并用于训练人工智能机器学习。笔者既不是律师也不是出版商，我不会发表意见，但作为一名观察者和作家，我是觉得这里面肯定什么地方有猫腻——而且看起来非常严重。

## **2\. Long- vs. short-term management**

2、长期管理与短期管理

Technology, by nature, forges ahead at its own speed. AI, being the most formidable technology ever, wants no reins, but that’s a recipe for nefarious actors – governments, corporations, and as yet undefined cyber-organizations – to use it for whatever they want. Follow that thinking to its foreseeable conclusion, and end-of-world scenarios are not out of the question, especially when aided by quantum computing, fusion energy, genetic engineering, biometrics, and facial recognition.

技术本质上是以自己的速度向前发展。人工智能是有史以来最强大的技术，自然也不需要约束，但这却也提供了政府、企业和一些尚未明确的网络组织等邪恶行为者利用人工智能为所欲为的可能。按照这种思路发展下去，世界末日的场景也并非不可能发生，尤其是借助量子计算、聚变能源、基因工程、生物识别和面部识别等技术。

But is this the best way to manage such a ubiquitous and so-far unruly creature? Enter, World Economic Forum.

但去管理这种无处不在、至今仍桀骜不驯的东西是最佳方式吗？世界经济论坛登场。

## **The World Economic Forum’s “Primer on AI”**

世界经济论坛举办“人工智能入门”研讨会

According to the WEF’s web site, the “Primer on AI” session will address the rapid growth of artificial intelligence, emphasizing “the critical need for inclusive, ethical, and well-aligned solutions” while exploring how to ensure AI benefits everyone.

根据世界经济论坛网站的介绍，“人工智能入门”研讨会将讨论人工智能的快速发展，重点放在“包容性、道德性和协调性解决方案的关键需求”上，同时将探讨如何确保人工智能造福于每个人。

Why? A quick history lesson might help.

为什么呢？一堂简短历史课或许会有所帮助。

When Johannes Gutenberg gave us the movable type printing press in in Mainz, Germany in 1450, his invention spread rapidly. By 1500, 1,000 presses across Europe had published 9,000,000 copies of more than 40,000 works. Just as nature abhors a vacuum, humanity abhors stagnation, and when Chat GPT launched late last year, it took only five days to cross the 1,000,000-user mark.

Johannes Gutenberg 曾于 1450 年在德国美因茨为西方带来了活字印刷术，他的发明得到迅速传播。到 1500 年，全欧洲的 1000 台印刷机共出版了 4 万多部作品的 9 百万册书。大自然痛恨真空，人类则痛恨停滞不前。Chatgpt 于2002年底推出时，仅用了五天时间就突破了 100 万用户大关。

That’s the thing. Each technology makes it easier to adopt the next one. With that, there’s growing concern that we better get it right on AI – now – or live with the consequences forever.

说的就是这个。每一种技术都会使得下一种技术更容易被采用。因此，越来越多的人担心，现在我们最好把握好人工智能，否则将承担永远的后果。

That’s how transformational AI is.

人工智能的影响之深远也正在于此。