I was talking with my teen this morning as we were getting ready for school and the subject of AI (in this case ChatGPT which is the favorite AI flavor among teens) came up. I don’t think we are talking about this enough. Lots of kids are outsourcing all of their thinking to AI. Not using it as a helpful assistant to find information or organize thoughts, but literally to do their “learning” for them. I find this terrifying.
We were talking about math as an example. Now, the way math is taught these days does have me looking for YouTube videos from helpful teachers that walk me through things I have forgotten or things that are just taught differently than they were in my education. But kids in general? They are literally just having ChatGPT do the work for them. YIKES.
I was explaining to her how AIs are programmed to sound confident even when they are making things up. And how the AI that is used by her peers is not thinking when asked questions, it is instead using a very sophisticated algorithm to guess what the next words should be based on its training data. It is not a true answer machine. It is a return something that sounds reasonable based on training data and a matching algorithm machine. But kids are using it in place of actually trying to learn things.
And look, I am not so old that I don’t remember wanting to find short cuts while in school. I get the urge. But we are about to potentially have an entire generation that literally cannot think or reason or problem solve for themselves. Who need an AI to do basic math and writing.
AI also is wrong, a lot. There have been numerous papers from highly respected institutions published recently about just how wrong the current LLM AI models are. Even some that show, conclusively, that this is part of how they have been built and it will only get worse as the models absorb and train on AI generated content.
I shared with her how dangerous it is for these AIs to sound so confident. Because if you’re asking an AI about something, chances are decent it is because the subject you’re asking about is not one where you are an expert. And that is where things get messy. Because how would you know if what it tells you is wrong? Especially when it sounds to confident in its answers? The short answer is – you wouldn’t. But we seem to be just ignoring that piece generally and taking what an AI spits out as accurate or correct.
I know the AI hype is massive. The global markets have decided that the AI industry is our future whether we like it or not. So there is a lot of pressure to utilize AI in many aspects of our lives. But before you just blindly take what AI generates as gospel, maybe double check it with someone who would know if it was wrong?
If you would like an expert set of eyes and an experienced PPC brain to review AI generated PPC strategy, I offer that service! I realize people are intrigued by the promise of AI, but the thought of them putting all of their eggs in an AI generated basket worries me.
If you want an PPC AI Strategy Double Check, get in touch here.
Leave a Reply