Stop Doing Easy Things

In 2011, Rich Hickey distinguished "easy" tasks from "simple" ones, defining easy as approachable and effortless while simple requires deliberate effort to achieve clarity. The rise of generative AI has made easy tasks even easier, inflating productivity metrics without delivering meaningful results. Developers should stop prioritizing easy, repetitive work that anyone can do and instead focus on creating simple, valuable solutions.

Stop doing easy things In 2011 Rich Hickey in his seminal talk "Simple made easy" defined the distinction between Easy and Simple. Easy is something that is approachable, can be done quickly and without effort. Simple, on the other hand, is something that is not convoluted, something clean and straightforward. It's sometimes very difficult to make something simple. The raise of generative AI pushed this concept even further. LLMs make easy things even easier, but they rarely produce simplicity. Doing many easy things quickly can inflate 'productivity' metrics and give a dopamine boost, but it's unlikely to deliver something amazing. Easy by definition is something that anyone can do. Easy leads to repetitiveness, duplication, waste. Stop doing easy things. Examples of easy things: - webpage, dashboard visualizing data without drawing any insights or developing understanding of the underlying patterns. - command line interface or MCP to an already existing service, without a good idea how it can be utilized by agents - yet another troubleshooting agent, with only simple prompt, access to Jira and only superficial understanding of the product - slack bot that summarizes channel content only to produce more content with reduced resolution - todo list app, without a clear differentiation