请问HN：当能力与资历脱钩时，会发生什么？
在过去的18个月里，我与人工智能合作，构建技术系统并进行分析工作，这些工作远远超出了我的正式培训范围。我没有计算机科学学位，也没有我所从事领域的背景，更没有任何机构的隶属关系。
这项工作是严谨的。某位拥有严肃资质的人参与其中，并提出了实质性的问题。这些系统按设计运作。然而，我无法指向那些能确立合法性的传统标志——学位、出版物、在该领域的工作年限。
这并不是在讨论人工智能“是否完成了工作”。我做出了每一个决策，评估了每一个输出，经过数百次的迭代精炼。人工智能只是一个工具，它将本该花费数年正式教育的内容压缩成几个月的集中、针对性的学习和执行。
我感兴趣的是：我们正进入一个传统能力信号——资质、机构验证、经验标志——不再可靠地预测能力的时期。现在，有人可以在没有任何历史上表明这些能力的资质的情况下，构建复杂的系统，进行严谨的分析，并产生新颖的见解。“能做”和“值得信任去做”之间的差距正在迅速扩大。
旧有的把关机制正在比新的机制形成得更快地崩溃。当资质不再是能力的可靠指标时，什么会取而代之？我们如何共同建立知识和能力的合法性？
这不仅仅是理论问题——它正在以规模化的方式发生。每天都有更多的人在构建事物和进行他们没有正式资格的工作，而其中一些工作确实很出色。
当传统信号变得过时时，我们应该使用什么框架来评估能力？当“专家”、“严谨”和“合格”等术语被稀释到失去区分力时，我们如何建立新的专业语言？
查看原文
Over the past 18 months, I've been collaborating with AI to build technical systems and conduct analytical work far outside my formal training. No CS degree, no background in the domains I'm working in, no institutional affiliation.<p>The work is rigorous. Someone with serious credentials has engaged and asked substantive questions. The systems function as designed. But I can't point to the traditional markers that would establish legitimacy—degrees, publications, years of experience in the field.<p>This isn't about whether AI "did the work." I made every decision, evaluated every output, iterated through hundreds of refinements. The AI was a tool that compressed what would have taken years of formal education into months of intensive, directed learning and execution.<p>Here's what interests me: We're entering a period where traditional signals of competence—credentials, institutional validation, experience markers—no longer reliably predict capability.
Someone can now build sophisticated systems, conduct rigorous analysis, and produce novel insights without any of the credentials that historically signaled those abilities. The gap between "can do" and "should be trusted to do" is widening rapidly.<p>The old gatekeeping mechanisms are breaking down faster than new ones are forming. When credentials stop being reliable indicators of competence, what replaces them? How do we collectively establish legitimacy for knowledge and capability?<p>This isn't just theoretical—it's happening right now, at scale. Every day, more people are building things and doing work they have no formal qualification to do. And some of that work is genuinely good.<p>What frameworks should we use to evaluate competence when the traditional signals are becoming obsolete? How do we establish new language around expertise when terms like "expert," "rigorous," and "qualified" have been so diluted they've lost discriminatory power?