我们需要真正关心的聪明人
We need smart people who give a shit

原始链接: https://violetpixel.com/blog/2026-11-17-we-need-smart-people-who-give-a-shit/

复杂、脆弱的系统,如ACH、银行和基础设施,之所以通常能正常运行,是因为“真正懂行且在乎的人”深入理解并手动维持它们。最近航空、银行、科技和政府系统中的广泛故障表明,这些人正在消失。作者认为,资本主义领导层被生成式AI的炒作和更低的成本所吸引,裁掉或掏空了这些专家,代之以模仿智能但缺乏真正理解、责任感和处理边缘情况能力的AI。迫使剩余员工照看LLM会进一步侵蚀他们的专业知识和主人翁意识。结果是:系统变得更加脆弱,故障增多。资本主义在被警告风险后,如今却在触碰热炉——而每个人都在被烫伤。

一个 Hacker News 讨论串讨论了一篇主张我们需要聪明且真正关心的人的帖子。评论者普遍同意,但表示关心的员工常因奖励速度、可见度和成本削减而非质量的企业文化而受到惩罚、边缘化或倦怠。许多人指责管理、裁员、OKR 和短期激励将员工视为可替代品,并抑制深度关心。其他人讨论 AI:它提升产出并帮助编程,但可能因降低对系统的理解而制造“理解债务”;一些人反驳 AI 直接导致了最近的停机。一些人指出,关心型工作得不到回报,可能拖慢速度,并可能威胁管理者,而“聪明且能完成任务”比长期维护更受奖励。总体而言,讨论表明当组织低估持续的关心、专业知识和问责时,技术系统会受损,而 AI 放大了速度却未匹配理解。
相关文章

原文

Something I learned a long time ago, something that's been proven to me time and again, is that the world is held together by smart people who give a shit. This truth always comes into stark focus when I learn how a complex system functions—or, rather, how it barely functions.

When I worked at Stripe I learned a lot about various banking and money movement systems, and it's terrifying how flawed and fragile these systems are. ACH is a great example. It's a complex system we rely on every day that is just barely good enough to function. It's old, riddled with flaws, and encumbered with legacy baggage. And yet, despite all of those problems, ACH usually works great and is generally considered to be safe, cheap, and reliable.

The reason there's a mismatch between the technical reality of the ACH system and how it functions in practice is due to a single factor: smart people who give a shit. Every problem with ACH is fixed, patched, bypassed, overridden, routed around, or manually handled by people who have two things:

  1. Intimate knowledge of the system
  2. The drive and determination to force the system to work

Those two things feed into each other. You have to care about a complex system to dive deep enough to learn how it works, and when you know a lot about how a complex system works you care more about it, especially if the system is important or critical infrastructure.

Now, you can replace "ACH" above with a myriad of other complex systems—everything from huge infrastructure to nice-to-have apps—and everything I said remains true, both about the systems themselves and the people who keep them running. The world runs on these systems, and they depend on smart people who give a shit. Take those people away, and the systems collapse.

And collapse is what we're starting to see. Some recent examples of various failures from just this past week or so: Microsoft 365, Air Traffic Control, United Airlines' systems, the NSA's network, many major banks, Apple's systems, and more. We're told these failures and problems are not related. They're not the result of a cyberattack. They're not the fault of generative AI.

I guess that depends on how you look at it.

I don't have inside information about what's going on, but my experience dealing with complex systems tells me there's a common through line to all of these disruptions: there aren't enough smart people who give a shit.

So where did they go?

Capitalism has never been particularly kind to smart people who give a shit (or people in general... but I digress). Capitalism does, however begrudgingly, recognize that smart people who give a shit are required to build and maintain complex systems that serve capitalist goals. Thus, those people have been accommodated and incorporated into the system by virtue of their necessity and, in many cases, their rarity.

Now, though, generative AI is here. To the people running the show—leadership and management—it seems like they can get the same or better value (and even more output!) from a machine. The machine is cheaper, works longer, doesn't take sick days or vacations, doesn't ask for any benefits, and never pushes back against a bad idea.

People who don't do the actual work—the people who aren't in the trenches—have a hard time telling the difference between a flawed system held aloft by the drive and determination of smart people who give a shit and slop provided by generative AI. To them it's all just output: words you can read, code that runs, and pixels that look nice to their eye. It's all the same to them, so why does it matter where it comes from?

It matters because generative AI isn't smart, and it doesn't give a shit. What seems like intelligence is an illusion propped up by a massive amount of data strip-mined from the commons with weights that nudge the output toward a generally desirable shape. What seems like care, concern, and attention is just generative AI doing what it's designed to do: telling you what you want to hear. The math and logic that power LLMs are as cold, unfeeling, and indifferent as the laws of physics.

Smart people who give a shit notice things. They connect dots. They notice when things don't feel right. They know when something is about to go off the rails, and they know how to get things back on track. They know how to handle edge cases. They know how to manually handle a unique situation. They know what the right thing to do is, and they know when they need to put in extra effort or time to make something happen. Generative AI falls far short in these areas, but this is the glue that holds complex systems together. More prompting and more output isn't a substitute for this.

It's hard for capitalist leadership to internalize any of that, though. It would be such a relief for them, such a burden lifted, to replace all those smart people who give a shit with generative AI. It would be so much easier to win at capitalism if generative AI lived up to the hype, and there's so much pressure for leadership and management to believe that the hype is real.

Thus, a lot of the smart people who give a shit have been laid off, and the ones that remain are hollowed out. Telling people to stop doing the job they signed up for in favor of babysitting LLMs so they can do that job instead is a great way to make people less smart about the system they used to work on and give less of a shit about it. Every prompt moves them one step further away from the system they used to know inside and out. Every generated output that becomes part of the system means those people understand less about how the system actually works. Those people are robbed of the things that make them smart (the struggle of building and learning) and the things that made them gave a shit (a sense of ownership and close connection with the system).

We've moved from a world dependent on complex and fragile systems held together by smart people who give a shit to a world of complex and fragile systems run by generative AI. The people running the generative AI are not as smart, nor do they give as much of a shit. There are, of course, exceptions (which is why a lot of things are still humming along), but we are on a course where the number of smart people who give a shit are decreasing, and that's a big problem. As long as the number continues to go down, more systems will fail.

I think what we have here is a hot stove moment for capitalism. When you tell a young kid a stove is hot and they shouldn't touch it, you'll find there are two kinds of kids. For one kind of kid, the warning is the lesson, and they don't touch the stove. For the other kind of kid, the only lesson that works for them is being burned. They have to touch the stove at least once to learn the lesson. Capitalism was told the stove was hot, but their hand is on it right now.

Unfortunately, we're all getting burned.

联系我们 contact @ memedata.com