AI是一种颠覆性的变革技术

Hello everyone, great to see you again!

你好呀，又见面了。

Learn English, Master AI.

学好英语，玩转智能。

Artificial intelligence is a transformative technology. It is the first technology to shorten its spread time from decades to just months. It turns cognitive work into large-scale industrial work, and puts this powerful change-making ability in the hands of a few global actors.

人工智能是一项具有变革性意义的技术， 它首次将技术普及周期从数十年缩短至数月，实现了认知工作的规模化、工业化， 并将变革性能力集中于少数全球主导力量手中。

More and more people view AI systems as a general-purpose technology. It can be used in as many fields as the steam engine, electricity, and the Internet.

越来越多的人将人工智能系统视为通用技术，其可应用领域之广泛，堪比蒸汽机、电力与互联网。

Its spread speed is faster than any technology before. It took decades for electricity to reach most households. It took about 15 years for the Internet to spread across the world and reach one billion users through the World Wide Web. ChatGPT reached 100 million users in only two months.

它的普及速度前所未有。电力走进大多数家庭花了数十年时间；互联网通过万维网实现全球普及、达到十亿用户规模，用了约 15 年；而 ChatGPT 仅用两个月就达到了 1 亿用户量。

Modern AI, especially foundation models, shows unprecedented economies of scale. This creates strong pressure towards centralization of capability. The most powerful AI systems need huge computational resources for training. Only a handful of global actors can access these resources.

现代人工智能，尤其是基础模型，展现出前所未有的规模效应，这给技术能力的中心化集中带来了强大推力。性能最顶尖的人工智能系统需要海量算力资源来完成训练，只有少数全球寡头能够获取这些资源。

AI has a huge impact on knowledge work. Earlier waves of automation changed physical labor. AI is the first to massively affect cognitive and creative work, including writing, programming, legal analysis, medical diagnostics, scientific discovery, forecasting and image generation.

AI对知识型工作产生了深远影响 。早期的自动化浪潮，改变了体力劳动，人工智能则是首个大规模影响认知与创造性工作的技术， 涵盖写作 、编程 、法律分析 、医学诊断 、科学发现 、预测以及图像生成等领域。

Large language models may present hallucinations as if they are real facts. Users often think confident-sounding language means the content is true and reliable. This disconnect between apparent competence and accuracy shapes the risk landscape in systematic ways.

大型语言模型可能将幻觉内容当作事实呈现。用户往往会认为，听起来笃定的表述，就代表内容真实可信。这种人工智能表面能力与实际准确性之间的落差，会以系统化的方式塑造风险环境。

葛峪精编：United Nations. Preliminary Report of the Independent International Scientific Panel on AI.