Raw LLM Responses

There's a HUGE problem with AI that I have yet to see being discussed, liability. If I ship crappy code/configuration/architecture I can be held accountable, even fired. You can't fire an AI agent. Leaders need someone to blame when things go wrong, or at the very least "here's where we went wrong, and here's why", at least not in a way that VPs will want to hear. AI can't offer that. Let's say you work at a law firm and they give someones job to an AI and they do the cases and gets the case totally wrong? That's a lawsuit. Saying they're going to replace everyone's job with an AI that can't be held accountable is not going to happen. They may solve this with some kind of insurance, but at the end of the day, you have to know what you're doing so there will always be a human reviewing, and if they don't review properly, they get... held accountable. Now, what it will do is make those that use AI SUPER productive. It will/is making one good engineer into five. But, unless a company has a need of 5X productivity they will layoff or not fill roles that are vacant, and stop hiring except for strategic roles. Your best bet is to get good at using AI and become a superstar at work. And honestly, this is bitcoin circa 2015. Many people are too lazy or couldn't care less about using AI in their jobs. Or they use it lazily. If you master prompts for AI to get the best outcomes you become like a magician.