寻求对一个纯AI人才市场创意的反馈
在过去几个月里，我经常需要为不同的公司寻找人工智能专业人才，但我不断遇到同一个问题：“人工智能专家”这个词如今既广泛又模糊。这让我想起早期互联网时代，当时总有一个“电脑达人”可以做所有事情。
公司说他们需要一个“人工智能人才”，但大多数时候，他们并不真正清楚自己在寻找什么样的角色。而候选人的技能各不相同，却都被贴上了同样的标签。我认为，在未来几年，人工智能就业市场将会爆炸式增长，角色会变得更加细分（如大语言模型工程师、机器学习运维、人工智能产品经理、应用研究员等），就像15到20年前的网络职业一样。
我的想法是建立一个专门为人工智能专业人士服务的门户网站：经过验证的个人资料、结构化的技能、项目经验、使用的工具、开源贡献、可选的薪资范围。没有信息流，没有社交功能，没有噪音。只提供人才与公司之间的对接。
公司可以使用真实的筛选条件（技能、技术栈、经验、薪资范围）进行搜索，并直接联系感兴趣的候选人，提供工作机会或项目提案——如果他们真的想要某个人，甚至可以进行“竞标”。
长期目标是创建一个生态系统，让人工智能专家能够接收来自多家公司的报价，进行比较，并朝着更具挑战性或薪资更高的角色发展。
最小可行产品（MVP）将会非常简单：个人资料 → 搜索 → 联系。没有复杂的人工智能功能，没有多余的内容。
我想问你几个问题：
你认为这个想法有意义吗？
从你的角度来看，这也是一个真实的问题吗？
我遗漏了什么或低估了什么？
任何反馈（技术、市场或产品方面）都非常感谢。
查看原文
In the past months I’ve often had to look for AI professionals for different companies, and I kept running into the same problem: “AI expert” today means everything and nothing. It feels a bit like the early Internet era, when there was the one “computer guy” who did literally everything.<p>Companies say they need an “AI person,” but most of the time they don’t really know which role they’re actually looking for. And candidates have very different skill sets but all get labeled the same way. I think that in the next few years the AI job market will explode and roles will fragment a lot (LLM engineer, MLOps, AI product, applied researcher, etc.), just like web professions did 15–20 years ago.<p>My idea is to build a dedicated portal only for AI professionals: verified profiles, structured skills, projects, tools used, open-source contributions, optional salary range. No feed, no social, no noise. Just talent → companies.<p>Companies could search with real filters (skills, stack, experience, salary range) and directly contact profiles with job offers or project proposals — even “bidding up” if they really want someone.<p>The long-term goal is to create an ecosystem where AI experts can receive offers from multiple companies, compare them, and move toward roles that are more stimulating or better paid.<p>The MVP would be extremely simple: profile → search → contact. No fancy AI features, no unnecessary stuff.<p>My questions for you:
Do you think this makes sense?
Is this a real problem from your perspective too?
What am I missing or underestimating?<p>Any feedback (technical, market, or product) is very appreciated.