AI Unethical Use Of

Ya know guys. I just started to mess around with AI the other day. So I started with something I knew. Pickups. You know how I knew it would take me an hour to do what Copilot did in seconds. I’d already been doing the research the old fashion way. So when it came back with answers I’d already found…. I thought that was kinda cool.

I worked behind a parts counter for 40 years. I had a saying that I never asked a question I didn’t already know the answer to. Always get a strange look. But when you’re dealing with the public…… farmer comes in and tells me his tractor doing this and he needs a widget. Experience tells me he really needs a thingy. But if I try and sell him the thingy he tells me wrong part. So I sell him the widget he wants. Then comes back and accuses me of selling him the wrong part and he walks back out of the store with the correct part. The thingy I tried to sell him before. So I learned to ask questions to help direct the person in the correct direction. Already knowing the answer.

That’s what I did with Copilot yesterday. Twas an experiment. One…. For myself to learn something about AI. Two…. See how accurate having, heard the stories. And just thought I’d share the experience.

Hindsight suggests I should have just kept it to myself.
 
I once asked ChatGPT, as a test, to describe to me an episode of Seinfeld I had just watched. It completely fabricated a story, suggesting situations that never happened on the show.

So I told it that it was wrong. And it said, “I’m sorry, you’re right, I was totally wrong.” And then it came up with a whole new episode, also comprising situations that never occurred on the show, let alone the episode in question.

So again I said it was utterly wrong. And again it said, “You’re right, sorry, I meant to say this.” And it spit out a third wrong answer, conflating bits and pieces of its first two concoctions into a third wrong answer. I could have gotten the right answer n five seconds from IMDB.

AI (which it isn’t) is a bad joke being played on software developers. Yesterday, my front door cam informed me that it now had AI… and would provide me with a written description of the video it had just recorded. WTF??!? How about I just watch the goddamn video?? So-called AI is performing functions no one needs. (It should without saying that small LLMs that only search a limited database, like in computer-assisted medical diagnosis, work great. Big LLMs Ike Gemini are useless and cost so much to run it’s ridiculous.)
 
iblive said:
Ya know guys. I just started to mess around with AI the other day. So I started with something I knew. Pickups. You know how I knew it would take me an hour to do what Copilot did in seconds. I’d already been doing the research the old fashion way. So when it came back with answers I’d already found…. I thought that was kinda cool.

I worked behind a parts counter for 40 years. I had a saying that I never asked a question I didn’t already know the answer to. Always get a strange look. But when you’re dealing with the public…… farmer comes in and tells me his tractor doing this and he needs a widget. Experience tells me he really needs a thingy. But if I try and sell him the thingy he tells me wrong part. So I sell him the widget he wants. Then comes back and accuses me of selling him the wrong part and he walks back out of the store with the correct part. The thingy I tried to sell him before. So I learned to ask questions to help direct the person in the correct direction. Already knowing the answer.

That’s what I did with Copilot yesterday. Twas an experiment. One…. For myself to learn something about AI. Two…. See how accurate having, heard the stories. And just thought I’d share the experience.

Hindsight suggests I should have just kept it to myself.
Click to expand...
Well, no need to be ashamed of sharing your experience. My post was about expanding the parameters of your experiment, finding out what it can/will do, or what it won't/can't.

I recently used it to help me troubleshoot an electrical issue with the 12v system in my RV, from my couch. It worked great, found all my relevant information for me, and I didn't have to go digging through diagrams and manuals (or the walls) as it just did all that for me. Just like copilot told you it can do.

It's a fine tool. Use it. Experiment with it. Play with it. It can be useful and it can save us time with certain tasks. That's all great and dandy.

We just have to remember how it arrives at it's answers and take that into account when we ask it things. If it's crawling through stuff like twitter/X/reddit posts for information, it can often come back with incorrect information, or even "hallucinate"

So yeah, like you alluded to, just remember not ask it anything that you don't already know, or know how to find, the correct answer to and assume it's the gospel truth.

While I don't love AI, I do understand that it's here and it's not going anywhere, it will become more and more ingrained into our everyday lives whether we like it or not, likely faster than we want. With that in mind, I feel like it's probably worth it to know it's capabilities and limitations. So continue the experiments!
 
mcblink said:
Well, no need to be ashamed of sharing your experience. My post was about expanding the parameters of your experiment, finding out what it can/will do, or what it won't/can't.

I recently used it to help me troubleshoot an electrical issue with the 12v system in my RV, from my couch. It worked great, found all my relevant information for me, and I didn't have to go digging through diagrams and manuals (or the walls) as it just did all that for me. Just like copilot told you it can do.

It's a fine tool. Use it. Experiment with it. Play with it. It can be useful and it can save us time with certain tasks. That's all great and dandy.

We just have to remember how it arrives at it's answers and take that into account when we ask it things. If it's crawling through stuff like twitter/X/reddit posts for information, it can often come back with incorrect information, or even "hallucinate"

So yeah, like you alluded to, just remember not ask it anything that you don't already know, or know how to find, the correct answer to and assume it's the gospel truth.

While I don't love AI, I do understand that it's here and it's not going anywhere, it will become more and more ingrained into our everyday lives whether we like it or not, likely faster than we want. With that in mind, I feel like it's probably worth it to know it's capabilities and limitations. So continue the experiments!
Click to expand...
I did ask Copilot if it had a thing going with Siri. That was entertaining...... and then it said, "shall we get back to taking about pickups?" :pound-hand:
 
You must log in or register to reply here.
Back
Top