5 Ways to Future-Proof Your Career in the Age of AI

自去年秋天以来，人工智能——特别是ChatGPT以及最近公之于众的其继任者GPT-4——已经在涉及未来商业、工作和学习的交谈中占据了中心位置。ChatGPT成为了历史上增长最快的消费者应用程序——超过了Instagram甚至抖音——而谷歌在一次拙劣的AI产品演示引发了对其竞争能力的质疑后，损失了1000亿美元的市值。

Since last fall, artificial intelligence — in particular ChatGPT and its recently unveiled successor, GPT-4 — has taken center stage in conversations about the future of business, work, and learning. ChatGPT became the fastest-growing consumer application in history — outpacing Instagram and even TikTok — and Google lost $100 billion in market cap after a botched AI product demonstration raised questions about its ability to compete.

这不仅对企业利益攸关，而且对希望驾驭自己职业生涯中的潜在巨大影响的个人而言也同样如此。多年来，专家们向我们保证，在AI的颠覆中得以幸存的关键是依靠创造力和其他（理论上）独树一帜的人类技能组合。但这一波新的AI正迅速证明，它能做的远不止处理数字和分析数据：它还可以创造专业水平的艺术和设计（参见DALL-E和Midjourney等），并创作文章和文案，可以取代许多记者和营销人员。

The stakes are high, not just for companies, but also for individuals hoping to navigate potentially massive implications for their own careers. For years, pundits have assured us that the key to surviving AI disruption was leaning into creativity and other (theoretically) unique human skill sets. But the new wave of AI is rapidly demonstrating that it can do far more than crunch numbers and analyze data: It can also create professional-level art and design (see DALL-E and Midjourney, among others) and create articles and copywriting that could displace many journalists and marketers.

我们相信，争论这个工具是否比我们“更聪明”或“更好”已经无济于事。这些工具就在眼前，而且它们已经得到广泛使用。对我们——托马斯，新书《我、人类：AI、自动化和重新寻回我们的独到之处》（I, Human: AI, Automation, and the Quest to Reclaim What Makes Us Unique）的作者，以及多里，个人品牌和职业发展顾问和主旨演讲人——来说，重要的是如何通过使用它们来提升自己。

We believe it’s no longer useful to debate whether this tool is “smarter “or “better” than us. The tools are here, and they’re already being widely used. What matters to us — Tomas, the author of the new book I, Human: AI, Automation, and the Quest to Reclaim What Makes Us Unique, and Dorie, a consultant and keynote speaker on personal branding and career development — is how we can better ourselves by using them.

当然，AI的战术应用几乎无穷无尽。比如，通过使用GPT-4或其他工具[例如微软现在集成了GPT-4的必应（Bing），或者尽管进展顺利，但最近又推出的Copilot]来协助进行集思广益，创建电子邮件的草稿，或进行快速复杂的研究，我们可以加快我们的工作效率。

Of course, there are nearly infinite tactical applications for AI. For instance, we can speed up our productivity by using GPT-4 or other tools (such as Microsoft’s Bing, which now integrates GPT-4, or its recently launched Copilot for all things work) to assist with brainstorming, creating drafts of emails, or performing rapid and complex research.

不过，除具体的用例之外，我们有兴趣探讨这样一个问题：即使在AI开始展示其非凡的（并且呈指数级增长的）力量时，专业人士是否仍然可以部署一些策略来产生独特的价值。换句话说，我们个人可以做些什么来避免AI可能造成的工作取代，并在智能机器时代保护自己的未来？以下是我们认为特别关键的五项策略。

But above and beyond specific use cases, we’re interested in exploring the question of whether there might still be strategies that professionals can deploy to generate unique value, even as AI begins to showcase its prodigious (and exponentially increasing) power. In other words, what can we do personally to stave off the displacement that may happen as a result of AI and future-proof ourselves in the age of intelligent machines? Here are five strategies we find especially critical.

避免可预测性

Avoid predictability

重要的是要记住，AI并不会产生新的见解；它是一个预测引擎，仅仅会猜测下一个最有可能的词汇。在微观层面上，它很有帮助作用：“谢谢”一词后面确实通常是“你”字。可是，在宏观层面上，它的建议通常会同质化，而且它们仅与群众的智慧相当，而群众的智慧通常恰恰是智慧的对立面。用奥斯卡·王尔德（Oscar Wilde）的名言来说：“一切流行的东西都是错误的。”王尔德可能不会成为ChatGPT的重度用户。

It’s important to remember that AI isn’t generating new insights; it’s a prediction engine that merely guesses the most likely next word. At the micro-level, it’s helpful: “thank” is indeed often followed by “you.” But at a macro level, its suggestions tend to homogenize, and they’re only as good as the wisdom of the crowds, which is often the exact opposite of wisdom. In the famous words of Oscar Wilde, who would probably not have been a heavy user of ChatGPT: “Everything popular is wrong.”

然而，如果你使用得当，生成性AI的这个方面可以成为强大的工具。比如，如果你希望了解大多数人对某事的看法或感受，包括他们的偏见和误会，你可以使用GPT-4来获取这些常识——或者无知。可是，如果我们只是不加质疑地部署这些工具，它们的算法和鼓励可能会把我们变成更可预测的生物，最终听起来全都一样。

And yet, this aspect of generative AI can be a powerful tool if you use it right. For instance, if you want to understand how most people think or feel about something, including their prejudices and misconceptions, you can use GPT-4 to access this common knowledge — or ignorance. But if we simply deploy the tools without question, their algorithms and nudges may turn us into more predictable creatures who eventually all start to sound the same.

请看看我们每次让Gmail自动完成我们的搜索或电子邮件时的情形，我们放弃了一点原创性和独特性，将AI的预测变成一个自我应验的预言，令我们更容易被人意料。虽然GPT-4是一个用于构思和创建初稿的强大工具，但如果你希望脱颖而出，你与它的建议反其道而行之可能会得到更好的服务（在某些情况下），因为你是在挑战传统的智慧。正如一些企业已经认识到的那样，模板式的“企业用语”会令客户转身离开，当其他每一个人都转向AI时，我们可能会发现，听起来像我们自己会具有优势——拥抱我们自己的个性、机缘巧合和不可预测性会带来好处。

Consider that whenever we let Gmail autocomplete our searches or emails, we relinquish a bit of originality and uniqueness, turning AI’s prediction into a self-fulfilling prophecy that makes us more predictable. While GPT-4 is a powerful tool for ideation and initial drafts, if you want to stand out, you may be better served (in some cases) by doing the opposite of what it suggests, because you’ll be bucking conventional wisdom. Just as some corporations have recognized that boilerplate “corporate speak” turns off customers, we may discover the advantage of sounding like ourselves — and embracing our own personality, serendipity, and unpredictability — when everyone else is turning to AI.

磨练机器努力模仿的技能

Hone the skills that machines strive to emulate

毋庸置疑：GPT-4已被训练得恭敬有礼（尤其是既然以前的——呃嗯——缺陷已经暴露出来）。它的回复会表现出同理心（“很抱歉我的回答令你不快”），自我意识（“我只是一个AI模型，我的回答依据的是训练数据”），甚至还有创造性 [创作关于不平等的精妙俳句，模仿埃隆·马斯克（Elon Musk），以及用多种语言生成无限多的奶酪笑话]。

There’s no question: GPT-4 has been trained to be respectful and polite (especially now that previous — ahem — vulnerabilities have come to light). Its responses display empathy (“I am sorry my answer upset you”), self-awareness (“I’m just an AI model and my answers are based on training data”), and even creativity (producing ingenious haikus about inequality, impersonating Elon Musk, and generating an infinite number of cheese jokes in multiple languages).

可是——重申一下——这些回复是基于文本预测的，AI没有能力体验或展示人类版本的这些软技能。正如托马斯广泛论证的那样，人类生来就会对真正的情感做出反应——因此，切实懂得和关心他人的想法和感受、真正了解自己，并且有能力创造出机器无法做到的东西，是在AI时代让自己与众不同的一项重要策略。

But — to reiterate — those responses are based on text prediction, and AI is not capable of experiencing or displaying the human version of these soft skills. As Tomas has argued extensively, humans are wired to respond to genuine emotions — so actually knowing and caring about what others think and feel, truly understanding yourself, and being capable of creating something machines cannot is an essential strategy to set yourself apart in the age of AI.

加倍关注“现实世界”

Double down on “the real world”

GPT-4——就像一般的AI一样——被局限于数字世界中，栖息于由0和1组成的虚拟牢笼里。可悲的是，如今许多人类活动也是如此。不过，我们必须认识到，AI无法颠覆的一件事是我们与他人的模拟式、面对面联系。因此，划出时间保护这些联系十分重要。

GPT-4 — like AI in general — is confined to the digital world, inhabiting a virtual cage of 0s and 1s. Sadly, so are many human activities these days. But it’s essential to recognize that one thing AI can’t disrupt is our analog, in-person connections with others, so it’s important to carve out time and safeguard those.

正如哈佛大学教授阿瑟·布鲁克斯（Arthur C. Brooks）对这项研究所做的总结那样，“科技排挤了我们在现实生活中与他人的互动，这会降低我们的幸福感，因此在我们的生活中必须小心翼翼地加以管理。”

As Harvard Professor Arthur C. Brooks summarizes the research, “Technology that crowds out our real-life interaction with others will lower our well-being and thus must be managed with great care in our lives.”

疫情之前生活中的人为现象，比如与同事一起进餐、参加会议、主动与陌生人交谈等，可能看起来不那么迫切了，因为我们已经很久没有做这些事情。然而，它们代表着建立联系和获得见解的机会，而这根本不可能通过AI实现——因此，它们代表着我们仍然拥有的独特竞争优势。

Artifacts of pre-pandemic life like having meals with colleagues, attending conferences, and initiating conversation with a stranger may seem less pressing now that we’ve gone so long without them. But they represent an opportunity to build connections and gain insights that simply aren’t possible through AI — and thus, they represent a unique competitive advantage we still possess.

同样，原始研究——实际与人交谈并发现新的见解——变得至关重要，因为AI只能将过去的点和已经提供给它的信息联系起来。当你通过自己的生活经验或新的访谈与会话来挖掘网上（尚）没有的信息时，你就为文化对话增加了一些真正的新东西，而这通过AI是无法做到的。

In a similar vein, original research — actually talking to people and identifying new insights — becomes critical, because AI can only connect past dots and information it’s already been presented with. When you tap into information that isn’t (yet) online through your lived experience or novel interviews and conversations, you’re adding something genuinely new to the cultural conversation that wouldn’t be possible through AI.

开发你的个人品牌

Develop your personal brand

AI工具已经达到充分的质量，以至于它们很可能会大量消灭许多行业的中低端市场（比如，自由职业市场上的文案和设计师，或者那些与具有成本意识的客户打交道的人，这些客户渴望甩掉费用而青睐免费的选择）。

AI tools have reached sufficient quality that they may well decimate the lower and middle ends of the market in many professions (for instance, copywriters and designers on freelance marketplaces, or those who work with cost-conscious customers eager to shake off expenses in favor of a free option).

在某些情况下，AI在质量上甚至可能与最高层次的专业人士媲美——但几乎可以肯定的是，那些行业的领导者不会被取代，这是因为他们具有品牌实力。就像艺术界的买家会支付成倍的价格去购买“伦勃朗的真品”而非由他同时代名不见经传的画家创作的同样美丽的画作一样，企业领导者可能会继续支付溢价，与被视为“他们领域的顶尖人物”的人合作——部分程度上是在宣示质量，部分程度上是就他们合作的对象以及他们看重的东西发表品牌宣言。

In some cases, AI may even match the quality with professionals at the highest echelons — but it’s almost certain those industry leaders won’t get displaced, and it’s because of the strength of their brands. Just as art-world buyers will pay exponentially more for a “real Rembrandt” rather than an equally beautiful painting by one of his lesser-known contemporaries, corporate leaders will likely continue to pay a premium to work with people viewed as the “top of their field” — partly as a statement of quality, and partly as a brand statement about whom they associate with and what they value.

举个例子，即使本地的轮胎店或花店都可以使用AI来创建徽标，但只有那些具有眼光和真正有钱的人——人们就是这样想的——才负担得起使用各机构的精英群体。AI的兴起并没有改变树立品牌十分重要这一事实，这与人类本性息息相关。

As one example, even the local tire shop or florist can use AI to create a logo. But only those with discernment and real money — so the thinking goes — can afford to use an elite group of agencies. The rise of AI doesn’t change the fact, crucially tied to human nature, that branding matters.

培养专业知识

Cultivate expertise

GPT-4和其他AI技术是非凡的研究者，几乎可以顷刻间唤出一连串的事实。不幸的是，其中部分事实并不真实——迄今为止一直困扰着AI的一种“幻觉”形式。（事实上，一位读者联系了多里，询问他在何处可以找到多里发表在《哈佛商业评论》上的、ChatGPT曾引用过的一篇文章。不幸的是，这篇特定的文章并不存在）。

GPT-4 and other AI technologies are prodigious researchers that can summon a cavalcade of facts almost instantly. Unfortunately, some of those facts aren’t true — a form of the “hallucinations” that have, heretofore, plagued AI. (Indeed, one reader contacted Dorie asking where he could locate one of Dorie’s Harvard Business Review articles that ChatGPT had referenced. Unfortunately, this particular article didn’t exist.)

因此，虽然AI是一种价值超凡的工具，但我们不能总是相信它会提供准确的结果——至少在这个时候。这就是在你的领域发展被人认可的专业知识如此有价值的原因。即使AI可以执行“初稿”功能，它仍然必须由一个值得信赖的可靠人士进行复核。如果那人就是你，你就会继续成为被人物色的对象，因为你拥有审查AI回应的权威。

Thus, while AI is an extraordinarily valuable tool, it can’t always be trusted to deliver accurate results — at least at this point. That’s why it’s so valuable to develop recognized expertise in your field. Even if AI performs “first draft” functions, it still has to be double-checked by a trusted and reliable source. If that’s you, you’ll continue to be sought out because you have the authority to vet AI’s responses.

AI技术有能力改变我们的职业生活——也许就在不久的将来。通过遵循这些策略，我们相信你可以找到发现和提供独特价值的方法，哪怕在GPT-4和其他技术发展进步的时候。即使在不断变化的时代，这也是为你自己提供职业保险最清晰的途径。

AI technology has the power to transform our professional lives — perhaps in the very near future. By following these strategies, we believe you can find ways to identify and provide unique value, even as GPT-4 and other technologies advance. Even in changing times, that’s the clearest path to career insurance for yourself.

关键词：AI

多里·克拉克（Dorie Clark）、托马斯·查莫罗-普雷穆日奇（Tomas Chamorro-Premuzic）| 文  

多里·克拉克是一位营销策略师和主旨演讲人，任教于杜克大学富卡商学院（Duke University’s Fuqua School of Business），并被“全球最具影响力的50大管理思想家”（Thinkers50）排行榜评为世界50大商业思想家之一。托马斯·查莫罗-普雷穆日奇是万宝盛华集团（ManpowerGroup）的首席创新官，伦敦大学学院和哥伦比亚大学的商业心理学教授，deepersignals.com的联合创始人，以及哈佛大学创业金融实验室（Harvard’s Entrepreneurial Finance Lab）的研究助理。

时青靖 | 编辑