Ask HN：作为一名工程师，如何保持领先地位？
3 分•作者: spacebuffer•1 天前
你好，在过去的几个月里，我一直在深入思考我的软件工程师工作，以及在人工智能（LLMs）飞速发展的今天，我未来几年将如何为我工作的公司提供价值。
每次当我以为找到了能让我保持竞争力的优势时，人工智能似乎在几周/几个月内就能做得非常出色。我举几个例子：
当 GPT-4 非常流行时，我们看到了大量写得不好的、仅凭感觉编码的应用。我曾认为网络安全是一个很好的职业发展方向，因为总有一天需要有人来保护这些应用，以防它们被投入生产使用。但现在，我们看到新的模型正以惊人的速度发现安全漏洞。
最近，我不得不在工作中处理一些棘手的难题，需要交叉引用数据库和服务器中的不同日志，以便快速定位问题，最大限度地减少对生产环境的影响。我非常享受这个挑战，并认为这对我和我的雇主都很有价值。但最近我遇到一个人，他给了 Opus 只读权限访问他的 AWS 公司账户，结果 Opus 的诊断速度比他自己还要快（据他本人所说）。
当然，鉴于我的服务器托管在没有只读访问权限的平台上，我目前还不能让 AI 代理访问我的服务器。但我预见这种情况不会持续太久。
我想听听你对这种情况的看法。
查看原文
Hi, for the past few months I've been thinking a lot about my work as a software engineer and how I'll be able to provide benefits to companies I work for in the coming years, given the fast pace of progress of LLMs<p>Every time I think I found an edge that will help me stay I relevant, AIs seem to get really good at it in a few weeks/months. I'll give a few examples.<p>When gpt-4 was really popular we saw an increase of badly written vibe-coded apps, I thought cybersecurity was a good career path to persue since sometime will eventually need to secure these apps if they start being used in production. but now we're seeing new models find security vulnerabilities at an insane rate.<p>More recently I had to deal with a few difficult problems at work where I had to cross-reference different logs in the DB and the server to pinpoint the issue quickly so production wouldn't be impacted for very long. I really enjoyed that challenge and found it to be rewarding both for me and for my employer. but I recently came across someone who gave opus read only access to his AWS company and it diagnosed faster than he ever could have (according to the guy himself)<p>Sure I cannot give AI agents access to my server yet given that its hosted on platforms without a read-only access but I don't foresee this lasting much longer.<p>I wanted to know what your thoughts are on this situation.