人工智能内幕网（AI Insider）本周前往盐湖城参加了“重大编程重置”（The Great Coding Reset）活动，这是一场关于软件工程未来的现场活动。我们听到的内容对科技工作者来说既令人鼓舞又令人警惕。如果你想参加我们12月1日在拉斯维加斯举办的下一场活动，请在此注册。
AI Insider went to Salt Lake City this week for The Great Coding Reset, a live event about the future of software engineering. What we heard was both encouraging and alarming for tech workers. If you want to attend our next one in Vegas on Dec. 1, sign up here.
在帮助企业运行复杂计算系统的CIQ公司，据总裁比约恩·霍夫兰（Bjorn Hovland）透露，人工智能现在编写了90%以上的代码。你可能会认为这意味著需要更少的人类程序员。事实并非如此。“两三年前，我曾预计我们可能再也不需要雇佣其他工程师了，”霍夫兰对我的同事本·伯格曼（Ben Bergman）说。“我们发现的情况恰恰相反。我们现在招不够工程师。”
At CIQ, a company that helps businesses run complex computing systems, AI now writes more than 90% of the code, according to president Bjorn Hovland. You might think that means fewer human programmers are needed. Not quite. "Two or three years ago I expected that we might never have to hire another engineer," Hovland told my colleague Ben Bergman. "We found the exact opposite.
这里有个问题。CIQ想要经验丰富的程序员，而不是初级程序员。“我们几乎不雇佣初级开发者，”霍夫兰说，并补充道他认为该公司没有雇佣过任何经验少于5到10年的人。
We cannot hire enough engineers now." There's a catch. CIQ wants experienced programmers, not beginners.
为什么？人工智能可以生成成堆的代码，但仍然需要有人知道它是否可行。当你的客户包括大公司和政府时，这一点尤其重要。CIQ创始人格雷戈里·库尔策（Gregory Kurtzer）提供了一个展示人工智能强大功能的引人注目的例子。他说他用它在短短11天内构建了一个与Linux兼容的操作系统内核——即控制计算机工作方式的软件。“这应该让人感到有点害怕，”他说。
"We hire almost no junior devs," Hovland said, adding that he didn't think the company had hired anyone with less than five to 10 years of experience. Why? AI can produce mountains of code, but someone still needs to know whether it works. That's especially important when your customers include major companies and governments. CIQ founder Gregory Kurtzer offered a striking example of AI's power. He said he used it to build a Linux-compatible operating system kernel — software that controls how a computer works — in just 11 days.
库尔策认为，使用人工智能的经验丰富工程师可以极大地提高生产力。他还警告说，由于热爱自己编写代码而抗拒人工智能的程序员，面临著落后的风险。
"It should scare people a little bit," he said. Kurtzer believes experienced engineers using AI can become vastly more productive.
对科技工作者而言，这个信息是复杂的。人工智能正在消除一些编程任务，同时为深入理解软件工作原理的人创造更多的需求。
He also warned that programmers who resist AI, because they love writing code themselves, risk falling behind. The message for tech workers is complicated.
这超越了编程。无论你是在制作视频、编写软件还是从事其他知识性工作，有价值的技能越来越多地在于知道创造什么以及如何判断结果——而不是自己去做每一个步骤。
AI is eliminating some programming tasks while creating more demand for people who understand deeply how software works. And this goes beyond programming.
但这里有一个令人担忧的问题：如果公司停止雇佣初级程序员，下一代经验丰富的工程师将从哪里来？霍夫兰给年轻人的建议是：自己构建大量的软件项目，并了解系统底层的运作方式。这是个好建议。但进入软件工程的传统职业阶梯可能正变得越来越难攀登。
Whether you're making videos, writing software, or doing other knowledge work, the valuable skill is increasingly knowing what to create and how to judge the results — rather than doing every step yourself. But here's the troubling question: If companies stop hiring junior programmers, where will the next generation of experienced engineers come from? Hovland's advice to young people: Build lots of your own software projects and learn how systems work under the hood. Good advice.
在这里。通过电子邮件与我联系：abarr@businessinsider.com
But the traditional career ladder into software engineering may be getting harder to climb. here. Reach out to me via email at abarr@businessinsider.com