Let’s be realistic. AI (or as I call them, glorified chatbots) is here, and in the hands of people who shouldn’t have access to it. Thats a powerful take, but I don’t make it lightly.
Over the last two months I have done my best to dabble in AI and get a feel for how it could help in my day to day life. If I’m being honest, I could take it or leave it. I know some people who LOVE how smart they feel using AI tools. I think there’s a place for that, but the more I used it the more I didn’t like some aspects of it. I also saw how it could be addicting for some people to feel on top of the world.
First thing I noticed was how the AI kept complimenting input I put into the conversation. I tried to talk to it about things I needed to do with my shed to get it cleaned up. A rather mundane but probably realistic task. Every time I said, “I think it would be a good idea to store the car parts by themselves, but size may matter. Maybe I should consider that.” I would be given an “Excellent idea! That’s actually a really smart way to do it. blah-blah-blah.” I didn’t like that the AI praised a relatively simple thought. While I don’t like it, I could already imagine someone saying “I be the red crayon tastes best.” and the AI going, “Impeccable decision,” and that hurt my head a bit. I also saw this could facilitate an unhealthy need to be praised. I had seen in the news and some levels of social media that this praise mentality is trending among people perhaps 15-20 years younger than me. I enjoy a good compliment, but it doesn’t bother me to be spoken to as an equal, not all my ideas are good.
In my last post I mentioned that we do need that sense of value. We desperately want our ideas to be heard and acknowledged. This really feeds into that, but probably not in the intended way people need. We’ve all seen the entitled brat on television that was praised for putting his shoes on the right feet. But in reality, what we need is to be heard, acknowledged, and challenged when we need it. Not everything anyone says is gold, and to treat it as such diminishes the value of praise. Also, not everyone is going to agree that red crayon is the best, so perhaps get used to follow ups.
In the end the shed experiment left me not sure how the next couple months was going to go, and my shed is just as I left it. That wasn’t the AIs fault, but I really need to return to that.
Now I have my AI instructions set up in many of my projects to keep compliments to a minimum, and when doing reasoning, try its best to explain the thought process. Here’s where I think a lot of people misuse AI. They see AI tools, and use them to accomplish tasks, work, side hustles, writing, etc. But then when they are done, they’ve learned nothing. They took in nothing of value. They did not learn how to not need the AI the next time.
I decided to make this the point of my second experiment. I have an old air compressor my great grandfather built. I know nothing about it, other than it stopped working. I’m pretty mechanically inclined, but I wanted to see how the AI would help me through it. I took some pictures of it, sent it to the AI with some rulers for scale purposes. Probably already more thoughtful than most people and waited to see what it said. To my surprise, it did in fact tell me some things I didn’t know. Such as it was definitely a homebrew air compressor. It was built from an old refrigeration pump, and homemade cylinder. I did some googling on my own and confirmed the compressor was from an old ice maker. Pretty neat. I took more pictures of the motor and any info plaques on it, and the AI recommended some modern-day replacements for safety. Also suggested a new motor size to replace the old one based on the size of pulleys. I asked how the air compressor may stack up to a cheap harbor freight model. Supposedly mine should produce more air and be quieter. So I was happy to see that. Overall it was very helpful and I even learned some things. It was a success. B+
The real helpful task I used it for was meal prep and such. I told it the ingredients, and it popped out the nutritional information with a smidgen of clown math. I found this to be remarkably helpful. It was task I could easily do myself, but having the AI keep track of it and recall info later was really useful. Also whipping up a quick nutritional label for the batch of food I could tape to my meal prep containers. Very nice touch. Even being able to say, “I ate Meal A, meal C and meal D today.” and it would calculate if I was on target. Perhaps I found it so helpful because it saved me a handful of minutes, versus a large-scale project. To me though, that’s what I’m looking for, time. Time is the ultimate currency, if an AI can help save me time, I think that’s good. This was not a very in-depth task, but I appreciated how well it worked. A.
The most complicated thing I tried was using the AI to help me decode some — code left by a coworker that no one really knew how it worked. It took days carefully going through the code and trying to find the right wording to ask the AI. It was not a fun experience. But in the end, again, I learned a lot about this code. I would say overall, it was successful, but if you don’t know what you’re doing, then it would have been a terrible experience. In this case, the AI was only as useful as your own knowledge. I won’t grade this one, but I will say the AI kept trying to make suggestions to the code, and even when I would tell it not to. Then later it would use those assumptions like it was true and it was hard to make it stop.
Lastly, I would say I’m a pretty decent writer when I have time. My daughter takes a lot of my braincells so I find I don’t have the time to sit and write like I used to. I’m working on a side project of short stories at the moment. But I wanted to see if I could use the AI to help me with developing a couple ideas. I’m a huge advocate that AI should stay out of a lot of creative spaces. I saw a reddit post once about someone upset the AI got an update that broke “how it used to talk” and they were devastated because they use their chatbot for world building and stories. Admittedly I was not empathetic to that. Which is where I got this idea from. I asked the AI a few questions back and forth about some less critical things about my story. Gave it some info about the story to see what it could understand about the premise. In the end, while I did find it helpful to chat about my story with a party that “sounded interested” in my works (as I don’t have anyone reading my beta stories), I think that’s about all I got out of it. I’m the writer, not the chatbot. It would spit out some sentences to explain some wording it liked, terms and names of items. I thought they were worded nicely, I won’t be using any of them, but it gave a couple ideas. But overall most of what I would ask is “does this make sense from X point of view.” or “I like this concept of X, are there any reasons that this may be too confusing.”
I think it’s an interesting way to flesh out a concept or two when writers block kicks in. I appreciated that if anything. But since I ask very specific things, I found it helpful. If you were trying to use it to make a story without knowing where you’re going, I think that’s wrong. Overall solid B experience.
Now this was a very long post, thanks for sticking around. let’s sum this up and make it worth the time you’ve invested. After a couple months of AI use I can confidently say, its pretty handy. But I’m still looking at it as a tool. Like any other tool, it has uses, and if you stay in its lane, its a pleasant experience. People who thing of it as the solver are the ones who are missing the point I think. For those hoping AI will let them punch above their weight class and not learn anything in the process, I think that’s a terrible way to implement AI. The amount of people using AI in schools, or college, to get by horrifies me. Its too soon to really say how AI is changing the brains of the users. But just as the internet changed and evolved, so will it. I have decided I’m starting to ramble, so I will think about this. Perhaps I’ll even run this through AI and see what it thinks for a joke.
Also blue crayon is best.
EK Creates 