How to Monitor When AI Could Impact Your Job (Full Transcript)

[00:00:00] Speaker 1: What are some of the biggest advancements you've seen in AI that leads you to this conclusion that people are going to really start losing their jobs soon? I want to start by saying my goal with this article wasn't to terrify people. It wasn't to get people to freak out. It was just to start a conversation because this conversation needs to be had. Knowing is really important. Knowing gets you ahead. It may not affect you today. It may not affect you in a year or two years. But if you know that this is coming and you don't bury your heels in, you don't bury your head in the sand, you're going to be better positioned for when it does eventually happen. So that really was my goal here. What I saw was a wake-up moment for myself. I've seen these models get better and better every year. The AI models are getting better at code first. The model companies like OpenAI and Anthropic have focused on code before everything else because it helps them build the models for everything else. Over time, they've gotten better and better. And you can think of it like, let's say the model today is at level 50 on code. It's probably at level 20 on everything else. But that means next year, the model is going to be at level 50 on everything else and level 100 on code. Very recently, we had a new model from OpenAI that I tried and I had a very eye-opening experience with. It was the first time it felt like it was far better than me at pretty much everything I do from a technical perspective. I am not doing the hands-on coding work anymore. It is quite literally just saying, hey, what do you want done? And it does it for me end to end. One of the things that I actually give as advice to most of my friends who are outside of tech is I say, build your own personal benchmark, is what I call it. So take your job and take a few things that you think AI will never be able to do or it's very far away from. And build simple prompts for the AI to say, hey, try this. Put it in today. And it may not be great today. The output might not be anything that impresses you. But every few months, these AIs get better. I recommend trying it on the newest one every few months. And you'll have your own warning bell for when this is actually going to come and affect you. And again, just because it can do something doesn't mean it's going to take your job. You give the example of a nurse. I think that's a good one, right? It's not going to replace nurses anytime soon. There are jobs that are more affected. If it's just something in front of a computer and you're not interfacing with other people in person, you are likely to be affected first.