Ask HN：对于一家初创公司来说，什么才是有意义的追求？
1 分•作者: _yjlee•5 个月前
作为一家初创公司构建产品时，几乎不可避免地会遇到需要转型的时刻，无论是在假设阶段还是产品发布之后。但我们如何确保转型不仅仅是方向上的改变，而是真正推动团队和产品的增长呢？
我的结论是：我们需要明确定义我们想要专注的领域，并且只在这个领域内进行转型。
这是因为我相信，从每一次尝试中获得的见解、经验和人脉都应该直接为下一个挑战做出贡献。当然，也有很多团队通过其他方式取得了成功，但对我们来说，这种方法似乎是实现快速增长和真正进步的最有效途径。
这就是为什么我决定专注于“Agent”（智能体）。
我清楚地知道，有些人会质疑“智能体只是炒作”或者“热度已经过去了”。但市场仍然关注着智能体，并且颠覆性的产品也在不断涌现。像 MCP 和 A2A 这样的协议也在发布，从而推动了智能体生态系统的快速增长。
与其争论这是一个红海还是蓝海，我将这种增长速度和生态系统发展视为一个不容忽视的机会——而这种快节奏发展所带来的不可避免的挑战，既是初创公司应该解决的问题，也是初创公司应该抓住的机会。
自 2023 年初以来，我一直在构建 LLM（大型语言模型）应用，甚至早于“智能体”概念成为主流。由于我现有的网络，我拥有数百个潜在用户的网络；这极大地影响了我专注于这个领域——智能体的决定。
现在轮到你了。你如何在自己的创业旅程中航行？我总是渴望向他人学习。
查看原文
When building a product as a startup, it’s almost inevitable that you’ll reach a point where you need to pivot, whether it’s at the hypothesis stage or after launching a product. But how can we ensure that a pivot becomes more than just a change in direction and actually leads to real growth for the team and product?<p>My conclusion was this: we need to clearly define the domain we want to focus on and pivot only within that space.<p>This is because I believe that the insights, experiences, and networks gained from each attempt should directly contribute to the next challenge. Of course, there are many teams that have succeeded in other ways, but for us, this approach felt like the most efficient path for fast growth and real progress.<p>That’s why I decided to focus on “Agent.”<p>I am well aware that there are skeptics who argue “Agents are just hype” or “The buzz is already over.” But the market is still paying attention to agents and game-changing products are continuing to emerge. Protocols like MCP and A2A are also being released leading to rapid growth of the agent ecosystem.<p>Rather than debating whether it’s a red ocean or blue ocean, I see this pace of growth and ecosystem development as an opportunity that can’t be disregarded - and the inevitable challenges that arise from this fast-pace development are both the problem startups are meant to solve and chances which startups should take.<p>I have been building LLM applications since early 2023, even before the concept of “agents” became mainstream. Thanks to my existing network, I have a network of hundreds of potential users; this greatly influenced my decision to concentrate on this field - agent.<p>Now it’s your turn. How are you navigating your own startup journey? I am always eager to learn from others.