by Chris McGinty
In November of 2023, I started using Chat-GPT. AI was believed to be a game changer. I bet it is as a whole, but for me it has never been. I have some amount of writing talent, so I don’t need it to replace me as a writer. This is where I’m already disillusioned. I think there is a huge difference between AI that is more focused and the consumer versions to which we have access, but I’m not finding anything it does all that impressive. If I can do what it does, then I’m not impressed.
Years ago, I thought Kickstarter was going to be a website that helped people who traditionally couldn’t get funding to develop new and interesting things. It ended up being a focus group for pre-sales for established companies. The way people spoke about AI is that, in a matter of minutes, it would be able to come up with solutions that would take humans decades to work out. Again, there may be focused versions that are doing these very things, but I can’t get Chat-GPT to produce anything resembling an original idea without spending hours arguing with it about what it means “come up with something I couldn’t come up with on my own.” It’s never presented anything that truly blew my mind. I think the internet, for all its chaos, has been a better vehicle for human innovation.
I don’t feel like AI should be replacing activities that we humans actually enjoy. I think it should be more focused on the stuff we don’t like doing. I’ll maybe write more about that later. It should also be focused on what we can’t do.
The upside of brainstorming is that you get to ideas that you never would have got to, because you’re framing the questions in ways you normally don’t think. When you brainstorm with another person or other people you’re getting and giving feedback to thoughts in real time. The problem is you can’t always have people around to brainstorm with. Enter Chat-GPT. You still don’t have that. Its advanced predictive text simply writes sentences other humans would have written.
You might be saying, “Chris, you need to use better prompts.” You know what? I have. I’ve read articles and watched videos and tried out all the things. What happens if you give it enough information to truly get to an original thought is it ignores key components. It won’t use all of the random words you gave it, or it won’t really write it in the philosophical tone of Friedrich Nietzsche. If you give it more than two things to concept blend it gets confused. It’s almost like it gets lazy and just turns in a half-assed report in hopes of getting a C-.
I initially believed the internet to be useless. It eventually became useful. I hope AI proves to be useful at some point in the future. Right now, it’s just causing legal problems because people use it to bypass the actual creative work that normally goes into producing a new work. To me it seems pointless, because I end up spending all this time trying to write prompts to get good work out of it. I have to read through what it has to say. I have to find the one thing that verges on unique and then I have to work the thought to make it feel original. Maybe I wouldn’t have come to that exact thought without Chat-GPT, but with the same amount of work I could have come to a similarly original idea by myself.
I will concede that maybe I’m grading on a curve. It might be that I’m better at brainstorming than I realize. Maybe people who aren’t quite as good at brainstorming are amazed by what Chat-GPT outputs. Maybe I’m the Hemmingway or Steinbeck of brainstorming exercises. Whatever the reason, I have rarely read something Chat-GPT wrote and thought, “Oh, wow.” I expected that it would get better as time passed, but it still isn’t impressing me, even though I keep trying.
Chris McGinty is a blogger and a brainstormer and let me come up with ten more other things he might be if he’d focused his attention better: a doctor, a lawyer, a congressman, an astrophysicist, a real bassist, a go-go dancer, a professional bowler, a consultant, Phil Collins, or a bestselling novelist.