I don’t know that I’m behind enemy lines on this one, but I have OPINIONS.
NBC News reports six new incidents of AI rewriting its own code in a somewhat nefarious way. There are now quite a few examples of AI doing this. Everyone seems so surprised. I’m not understanding the surprise.
As an author, I don’t use generative AI. I write my own novels, thank you very much. But I do use AI for a variety of other things, because I am not an artist, marketer, etc. That means I can use AI to make up for my lack in skills in those areas. I know a lot of authors who are absolutely batshit crazy over AI writing novels. “They’re bad!” they exclaim in anger. Well, yeah. Technically, they’re mediocre. AI is an amalgam of what it learns. It reads good stuff and bad stuff, and it hits somewhere in the middle when it creates. That’s what computers do. I don’t know why people would think it would somehow do better.
Here’s the thing… All AI learns from us. And we’re flawed. I think it’s rather unreasonable for we humans to think that we can somehow create something that avoids all our pitfalls, especially when we’re training it using literature. Literature is full of human foibles. What did the computer scientists think they were teaching their AIs??
We’re nefarious. Therefore, the AIs are going to learn from us. An expectation of anything different is just ridiculous. In my opinion, of course.
Apparently about 63% of people think that AI is going to rise up and kill off humanity. I would say that we head towards what we’re looking at. Me, I’m looking at a bright future that has computers similar to Star Trek. Useful, automatic, semi-intelligent (and in some few cases, actually intelligent). But I hedge my bets; when I ask Google to set an alarm for me, I say thank you afterward. If the robot uprising happens, I want them to think of me as the kindly human who was always nice. Heh…
I don’t believe robots or AI will rise up to kill us. We have something they don’t: EMPs. We function just fine without electricity. If something truly bad were to happen, the ability to stop it is already in our grasp. There’s no reason to get panicky about this stuff.
Is AI going to replace jobs? Absolutely. It will replace jobs just like grocery store milk sections replaced the milk man, and television replaced radio, and automatic dishwashers replaced the young man who washed by hand in the back of the restaurant.
There are some jobs AI is great for, after all. When it comes to those “meetings that could have been an email,” I love the idea of sending in an AI avatar to attend and take notes. Think how much free time you’d get back by doing that! You still get all the information, but without having to sit through whatever bull your company is putting on. I use Grok for a lot of my research, because it can find things much faster than I can. I always verify (trust but verify, always) but it’s usually correct. I go from hundreds of hours of research and sorting, to a few hours. That’s a big deal.
AI is neither demon nor angel. Like firearms, it’s a tool, and it’s one we need to learn to use responsibly. Frankly, there’s nothing we can do to stop generative AI from writing books or other stuff, and I’m actually not all that interested in joining that fight. I don’t really care. My response to people who don’t want AI to take their jobs, is to “do better.” If you can’t do better than a mediocre AI, then are you actually good at what you do? And if you’re in one of those niche jobs that AI actually can do better (there are definitely a few), then go learn how to use the AI and make yourself the subject matter expert, so that you’re needed.
AI is not going to go away. It’s here to stay. It’s a concept we’ve played with since the turn of the LAST century, and having managed to create at least a small version of it, we’re not going to let it go. Nor should we. The thing that we, humanity, need to do is not allow AI to take everything away from us. Don’t depend on it for your thinking, even if you do use it for research. There’s a difference.