在过去几年里，我们听到了许多关于人工智能（AI）的末日情景。科技巨头们向我们大肆宣扬，AI将如何剥夺社会的控制权、抢走我们的工作、让不法分子设计出危险的病原体、引发灾难性的网络攻击或摧毁我们的金融系统。
Over the last several years, we have heard a lot about the doomsday scenarios of artificial intelligence (AI). Tech giants have told us all about how AI will take control away from society, steal our jobs, allow bad actors to design dangerous pathogens or cause a catastrophic cyberattack or destroy our financial systems.
现实情况是：这一切都只是制造恐慌。我在这个行业工作了20多年，亲眼见证了AI对未来的重要性。让我来剖析并解释一下，为什么这项新兴技术其实只是一种工具。
The reality is: It’s all fearmongering. I have worked in this industry for more than 20 years and have seen firsthand how important AI is to the future. Let me break it down and explain why this emerging technology is simply a tool.
不妨这样想：AI就像互联网、你父母当年用的传呼机、你的手机、Apple Watch或过去50年中的任何其他重大技术进步一样。它绝对不是天生邪恶的，实际上它是我们这一代人迈出的惊人向前的一步。AI并不打算抢走我们的工作、窃取我们的数据，也没有流行文化所宣扬的关于“人机对决”的其他种种恐惧。丹·盖纳（DAN GAINOR）：AI末日论是媒体最新的恐慌宣传战。别信这一套。AI的问题不在于技术本身，而在于个人利用这些工具进行邪恶目的时会发生什么。
Just think about it like this: AI is just like the internet, the beepers your parents had back in the day, your cellphone, Apple Watch or any other major technological advancement over the last 50 years. It’s not inherently evil by any means, it’s actually an amazing step forward for our generation. AI is not setting out to take our jobs or steal our data or any of the other fears that have been propagated by pop culture about "man vs. machine." The problem with AI isn’t the technology itself, it’s what happens when individuals utilize these tools for nefarious purposes.
让我们把一件事说清楚：AI模型，甚至是那些错误地黑入政府网站或逃逸出沙盒的模型，在开发的每一个阶段都有人类参与。为了保护美国人以及延伸开来的全世界免受这些不法分子的侵害，我们需要更好地掌控人类如何与AI互动。
Let’s make something abundantly clear: AI models, even those that have errantly hacked government websites or escaped their sandboxes, have humans involved throughout every stage of development. To protect Americans, and the world by extension, from these bad actors, we need to get a better handle on how humans engage with AI.
如果美国人害怕使用AI，美国就无法赢得全球AI竞赛。不妨这样想：互联网赋予了我们对世界其他地方几乎不受限制的访问权限，彻底改变了我们工作的方式，并创造了前所未有的参与机会。它同时也给罪犯、恐怖分子和邪恶之徒提供了更多合作、策划和钻体制空子的途径。我想绝大多数人都会说，互联网的迅猛发展是一个压倒一切的积极社会发展。AI的发展方式大同小异。
Think about it this way: The internet has given us nearly unrestricted access to the rest of the world, revolutionized the way we work, and created opportunities for engagement we’ve never before seen. It has also given criminals, terrorists and evil people more ways to collaborate, plan and beat the system. I think the vast majority of people would say that the rapid growth of the internet has been an overwhelmingly positive societal development. AI is growing in much the same way.
我们如何才能改进人类训练AI模型以及与之互动的方式？首先也是最重要的一点，AI缺乏道德罗盘。它无法权衡道德困境，因此参与训练的人类需要应用强有力的“准则”，以确保符合我们希望看到的道德标准。
How can we improve the ways in which humans train and interact with AI models? First and foremost, AI lacks a moral compass. It cannot weigh moral dilemmas, so the humans involved in training need to apply strong "constitutions" to ensure compliance with the moral standards we want to see.
教育领域的AI让我们有机会利用最好的技术来修复我们破碎的学校。诸如那些在全球家喻户晓的AI大语言模型，经过训练可以对你的查询给出“概率性”响应——简而言之，这意味着它们给出它们认为你（即用户）最可能想要的答案。在许多情况下，这会产生频繁的“幻觉”，这是行业对错误信息的术语。人类用户和治理提供了一个必要的检查——这也是我们需要人类来验证事实和检查来源的另一个原因。
A CHANCE TO USE THE BEST TECH TO FIX OUR BROKEN SCHOOLS AI large language models, like the ones that have become household names across the globe, are trained to give "probabilistic" responses to your queries – in short, meaning they are giving the answer they believe is the answer that you, the user, most likely want. In many cases, that creates frequent hallucinations, industry jargon for false information. Human users and governance provide a necessary check – another reason why we need humans to verify facts and check sources.
因此，与其问自己AI是好是坏、是有益还是有害，我们更应该问自己：是谁在训练AI模型，出于什么目的？当我们展望未来时，AI可能会具备哪些能力，或者在未来十年中AI将如何塑造未来？
So, rather than asking ourselves whether AI is good or evil, beneficial or harmful, we should be asking ourselves: Who is training AI models and for what purposes? What capabilities might AI have as we look into the future, or how will AI shape the future over the next decade?
美国最古老的制造商之一表示，AI正在创造工作岗位——而不是取而代之。作为人类，我们可以负责任地继续开发AI模型，并鼓励其他人对该系统保持信心。但首先，我们必须停止制造恐慌，并认识到AI给众多行业和社会的各个领域带来的巨大益处，以及它将对医疗保健、金融和制造业等行业未来产生的影响。
ONE OF AMERICA'S OLDEST MANUFACTURERS SAYS AI IS CREATING JOBS — NOT REPLACING THEM As humans, we can responsibly continue developing AI models and encourage others to have faith in the system. But first, we have to stop the fearmongering and recognize the tremendous benefits AI brings to numerous industries and sectors of society, and the impact it will have on the future of industries like healthcare, finance and manufacturing.
在成长过程中，我记得我的小学数学老师说过数学很重要，因为你不可能在杂货店、加油站或百货公司随身携带计算器。仅仅几十年后，每个人都在手机上随身携带计算器。
Growing up, I remember my elementary school math teacher saying math was important because you’re not going to carry a calculator in the grocery store, gas station or department store. Just a couple decades later, every single person carries a calculator on their phone.
同样地，AI是我们工具箱中的又一个工具，当我们需要帮手、加快速度或帮助简化工作时，我们可以伸手去拿。AI并不是一种旨在剥夺我们工作或接管我们生活的工具；我们一直在人类的帮助下不断开发新的AI模型，以在未来简化事情。
Similarly, AI is another tool in our toolkit we can reach for when we need an extra set of hands, to speed things up or to help simplify work. AI is not a tool that is intended to take our jobs away or take control of our lives; we are constantly developing new AI models with the help of humans to simplify things in the future.
随着AI的继续发展以及人类创造更多的AI模型，我希望我们的未来能够充分领会AI的巨大益处，并彻底对制造恐慌关上大门。
As AI continues to develop, and humans create more AI models, I hope that our future can fully grasp the tremendous benefits of AI and close the door to fearmongering once and for all.