You ask an LLM to code something and then just run the code blind without reading it and if it seems to work, publish the application.
You can imagine how many performance and security problems it produces.
I used to laugh at these, but I’ve just been asked to help with an issue where I work and what do you know… long story short it’s not funny to me anymore
You ask an LLM to code something and then just run the code blind without reading it and if it seems to work, publish the application. You can imagine how many performance and security problems it produces.
I used to laugh at these, but I’ve just been asked to help with an issue where I work and what do you know… long story short it’s not funny to me anymore