Feels like „AI agents” topic is all over the internet now (at least in my corner of the web/social media). Are they really anything useful or is it just another overhyped term?
As someone who has been building "agents" for about a year, definitely over-hyped. They work well in extremely narrow use cases where:<p>1. A human is there to keep it on the rails<p>2. Tools are domain specific<p>Most agents being shown are basically LLM calls encoded in a graph structure - which I don't think an agent is. As a part of having "agency", the agent needs to have control over its control flow, sans the start and stop node.<p>For example, coding agents are great because they operate in a highly predictable environment and can verify their outputs iteratively using a combination of static analysis and tests.<p>It's like when first self-driving cars came out. The real work is slow, painful, and involves good engineering.