▲ | ehnto 19 hours ago | |
> For example, I love Cursor. Especially for boilerplating. But SOTA models with tons of guidance can still not reliably implement features in my larger codebases within the timeframe it would take me to do it myself. Test-time compute and reasoning makes things even slower. Importantly it also takes you guiding it to complete the task. Meaning you still need to pay a human and the cost of the LLM, so it's slower and a bit more expensive. I am not convinced either that AI working on complex programming tasks could be guided by less skilled devs, meaning you still need to pay the skilled dev. In my experience so far, the cost analysis doesn't work for more complex application development. Even if the cost of the LLM was free it is often wasting the skilled dev's time. All these metrics will change over the years and maybe the math works out eventually, or in specific circumstances, and I forsee LLMs assisting in development into the future. I am not seeing the cataclysmic wholesale replacement of humans in the workforce some are predicting, at this stage. |