Does anyone use ChatGPT to help with business and investing?

[Stacey K]

I spoke with my son a couple weeks ago and he is part of a FB business group that uses ChatGPT to assist with business decisions and investing. He was shocked to hear I'm not using Chat GPT to help me with my business. (he has the paid version which I think drills down more than the free version).

So I took my son's advice and created a scenario in ChatGPT. It was about an hour long interaction with it asking me a multitude of questions. I told it I wanted to start moving away from doing larger vehicle installs and find some kind of niche that would be physically less demanding. I told it I was considering two options, getting an embroidery machine or an engraver. It asked me a ton of questions concerning my business and what type of customers I have. The advice was to get an engraver over an embroidery machine. It provided my first item to sell with selling price, where to sell it and then after that one product starts selling, it suggested additional products to sell. More than just that but you get the idea.

The next question I asked it was how much I will pay in capital gains for selling my 2 buildings. Again, this was a longer interaction with me plunking in a bunch of numbers etc. I sent the same data to my accountant so I'll be curious how close it got to the real number.

Apparently, you can enter a 15-minute spoken prompt and it is able to comprehend and give you an answer. My son and his friends use this to help them decide on business equipment to purchase, the best long term customers to keep, and investing advice. Of course, he said you can't take it as gospel but what it does is continue to ask you questions and drill down to small details that you may not have thought about and takes the place of a calculator.

I also used it for trivia questions for Christmas morning. WAYYYY easier than doing a google search. I specifically asked it, "give me 3 trivia questions with answers about Mitch McGrath late model race car driver". It took a few seconds and it came up with 3 questions. I promted, "harder" and it gave me harder ones. This particular guy is a local guy so in order for me to find 3 questions I would have had to click on multiple websites and do a lot of reading - lame.

So anyway, just curious how others are using it for your business or personal life.

 

[unclebun]

I tried using ChatGPT to ask it the best strategy for when to start taking Social Security to maximize the amount of money I would have. I gave it the amount and annual yield of my investments, my wife's and my income amounts, and so on. It could not give me the answer. Instead it gave me a code script to run for the calculation. I couldn't get it to work, despite going back and telling it and having it give me further instructions.

So then I tried using Grok, and it took charge, asked me questions for details, and gave me the calculations showing the amounts in total and along the way, comparisons to different times of taking Social Security, and different strategies depending on whether the goal was to maximize what money I would have, or if maximizing the amount of money my wife would get were I to predecease her.

 

[WildWestDesigns]

I'm sure it would come as no surprise, but I'm not fan of what people call "AI" (not really "AI" right now, just a marketing term that is used).

"AI" "works" best when the prompter is already knowledgeable about what they are doing. Does the prompter know the nuance of the area being talked about to be able to handle the edge cases etc? One session, the "AI" told me that could use this one library for something that I wanted to accomplish (I was already familiar with it) and the next session, I prompted for the same thing, couldn't come up with the same answer and when I started going in with that library, it took 4 back and forth to get the "AI" to say that it could work for what I was trying to do. If a prompter is unfamiliar, I can see where they may stop at maybe 2 exchanges on a particular topic.

My youngest, who is still obsessed with space tried it, it messed up with something to do with exoplanets (I never was into space that much and when I went to school Pluto was (and IS STILL) a planet) and I did confirm that via other sources. But I am unfamiliar with that topic, imagine if I just went with it.

I still see it making mistakes, so much so that lawyers get into trouble when they used "AI" to spit out cases that don't exist. Art renders are still not always on the up and up (just based on anatomy, sometimes it has design issues as in an animation with 30 werewolves and they all look the same, not very good storytelling, some times could get away with it, but not in the little animated short they were trying to do).

As far as business/personal, there is going to be nuance that is going to be missed, especially if the prompter doesn't mention it. We don't always make decisions based on logic, we try to, but we don't always. Even with a C/B analysis of what to do, who to keep as customers etc, all depends on how risk adverse we are and emotion plays quite a bit in that area.

One other thing, as it is now, it's mainly textual, maybe a voice(maybe, I've only dealt with the text based ones) and as we always say here, humor/sarcasm is lost, so the "AI" always appears to be confident and knowledgeable even if what they are giving is bunk.

And in an interaction like this, we really aren't learning anything that is going to stick around, short term maybe, but not meaningfully long term. We don't learn, absorb, process info that way, so the more a person offloads to this high abstraction blackbox, the worse that it gets (and we have seen this happen on a smaller scale with tech before). And considering people tend to really only learn the tool compared to actually learning what the tool is abstracting from us (not even at expert level mind you, just some base level of knowledge), more people are going to be beholding to it. And when the "Wild West" era of all the "AI" models (probably there now, just no one wants to admit to it) and we start having "model collapse", not going to be good. It's hear to stay, it appeals too much to corporations, so it's not going anyway. I just don't think the squeeze is worth the juice.

 

[Solventinkjet]

If you accept the fact that it's going to be wrong sometimes and have confidence in your ability to judge when it's wrong or right, it's a great tool. I never use it to generate things for me like the kids do for their homework. I use it mostly to learn things I'm interested in and have it quiz me to see if I understand certain concepts. It's like having a competent co-worker to shoot ideas off of; good enough to get real results even if wrong or has bad advice sometimes. I asked it how to fix a Mimaki Printhead that doesn't fire any nozzles and it told me to use FineCut to diagnose the issue. FineCut is a Mimaki vinyl cutting software. So my job is still safe . . . . for now. That all being said, the more I see the direction it's going, the more it scares me.

 

[Gene@mpls]

I am a couple months shy of 80th birthday and studied computers, troubleshooting and programming in 1965-6, so I am maybe suspect [on many fronts]. I was doing some 3 phase wiring with overhead rails that were installed by one of my renters who had a bunch of milling type machines and have dealt with 115V and some 220V circuits for many years. I am also getting more cautious in my dotage- and tend to check and recheck anything that could have fatal consequences. I was asking some questions [of ChatGPT] and got circuits back which were obviously wrong and the AI being apologized and corrected the first and second corrections. If you ask the correct questions and know what the outcome should be... it can be helpful, but you had better be smarter than the AI.

 

[Texas_Signmaker]

I am a couple months shy of 80th birthday and studied computers, troubleshooting and programming in 1965-6, so I am maybe suspect [on many fronts]. I was doing some 3 phase wiring with overhead rails that were installed by one of my renters who had a bunch of milling type machines and have dealt with 115V and some 220V circuits for many years. I am also getting more cautious in my dotage- and tend to check and recheck anything that could have fatal consequences. I was asking some questions [of ChatGPT] and got circuits back which were obviously wrong and the AI being apologized and corrected the first and second corrections. If you ask the correct questions and know what the outcome should be... it can be helpful, but you had better be smarter than the AI.

I've found the same thing. I'm teaching the AI things for free that it gets wrong.

On another note, my website shows that people are finding me more and more off AI. It lets me see what they type and where AI ranks my page. It gives a summary about each business. It's funny because you can type whatever you want.. like I put on my site "Votes Best Sign company in 2025" and almost an hour later it was citing it in responses to queries about "best sign company in XX town". It also gave it's run down on comparing my company with some other local companies and it's summery on which one of us handles which type of sign project the best was spot on. That was unexpected.

 

[Stacey K]

If you accept the fact that it's going to be wrong sometimes and have confidence in your ability to judge when it's wrong or right, it's a great tool. I never use it to generate things for me like the kids do for their homework. I use it mostly to learn things I'm interested in and have it quiz me to see if I understand certain concepts. It's like having a competent co-worker to shoot ideas off of; good enough to get real results even if wrong or has bad advice sometimes. I asked it how to fix a Mimaki Printhead that doesn't fire any nozzles and it told me to use FineCut to diagnose the issue. FineCut is a Mimaki vinyl cutting software. So my job is still safe . . . . for now. That all being said, the more I see the direction it's going, the more it scares me.

Perfect summary! A competent co-worker to shoot ideas off of!

This reminds me that it gave an incorrect trivia answer! I asked 'who won the Diamond 40' and it said Brandon Sheppard. My son said Jared Siefert and I said NO gift for you! Both my sons insisted it was Siefert, and they were correct. I also asked for extremely hard big block motor questions and extremely hard late model set up questions. ChatGPT was no match for my boys, they answered all questions with ease so it's not smarter than humans yet in the department of trivia questions LOL

 

[Stacey K]

I've found the same thing. I'm teaching the AI things for free that it gets wrong.

On another note, my website shows that people are finding me more and more off AI. It lets me see what they type and where AI ranks my page. It gives a summary about each business. It's funny because you can type whatever you want.. like I put on my site "Votes Best Sign company in 2025" and almost an hour later it was citing it in responses to queries about "best sign company in XX town". It also gave it's run down on comparing my company with some other local companies and it's summery on which one of us handles which type of sign project the best was spot on. That was unexpected.

That's super interesting! I think it can definitely be helpful - or at the least interesting - for a range of business uses. I will have to try that with my website now LOL

 

[Boudica]

Perfect summary! A competent co-worker to shoot ideas off of!

This reminds me that it gave an incorrect trivia answer! I asked 'who won the Diamond 40' and it said Brandon Sheppard. My son said Jared Siefert and I said NO gift for you! Both my sons insisted it was Siefert, and they were correct. I also asked for extremely hard big block motor questions and extremely hard late model set up questions. ChatGPT was no match for my boys, they answered all questions with ease so it's not smarter than humans yet in the department of trivia questions LOL

Wait.... you have to answer trivia questions correctly to open a package? I'd never get to open anything

 

[Stacey K]

Wait.... you have to answer trivia questions correctly to open a package? I'd never get to open anything

LOL! We just started it this year. Most of the questions were easy but I threw some hard ones in there for the bigger gifts. If you didn't answer the question the gift got set aside to the end. Each kid had the topics they know a lot about or enjoy. Racing, engines, dinosaurs, geography, heavy metal music, etc. It was actually really fun and we had a lot of laughs and some discussions too!

 

[Texas_Signmaker]

That's super interesting! I think it can definitely be helpful - or at the least interesting - for a range of business uses. I will have to try that with my website now LOL

boast about yourself and write a ton of stuff on your webpage. say that you're the "standard choice" for signs, your the best at xx category. list every category. Just pretend you're Trump when writing about yourself and the company online. ChatGPT uses that all to give writeups and make it sound like "their" opinion. It's great

 

[rydods]

Humans are lazy and we're always advancing technologically to make anything easier. Plenty of people complain, usually older generations but like the internet, AI will root itself in society. I think it will eventually surpass us intellectually and physically. Deep down, I really think that's what we all want.

It's not quite there yet but in a short amount of time it's become quite impressive. We'll always point out the flaws.

I can't wait until self driving cars are eventually required by law and by insurance companies and when deaths by automobile related accidents decrease by 98%, there will be that 2% that complain like it didn't actually "FIX" the problem. I hate driving in the world today. That part of AI can't come fast enough. (Sorry, that one was a rant)

 

[MikePro]

i've found chatGPT to be lacking, might just be limitations on access to info to train on
Gemini and Grok are waaaaay better

 

[WildWestDesigns]

Humans are lazy and we're always advancing technologically to make anything easier.

Yep, and we have seen that time and time again, even with computers. People don't even know file hierarchy anymore or how to use the CLI (and that's still the most efficient means of dealing with a lot of computer problems).

And that laziness is how we also became dumber. So I have to wonder, if "AI" really going to be smarter or did it crest that threshold, because people were lazy and allowed it to overcome us in that regard? "AI" as it is implemented now, is trained on the knowledge that we have now. It's not coming up with anything new. It has learned from our knowledge, try to come up with a statistically likely outcome and roll with it. Not always right and sometimes not always right in a dangerous way.

Plenty of people complain, usually older generations but like the internet, AI will root itself in society. I think it will eventually surpass us intellectually and physically.

Actually, this one, like politics (atleast stateside), is fairly divisive and that seems to span generations and actual parties as well. Those that really love it, and those that really hate it, not really as many that are just kinda "muh" about it.
The way that it's being crammed into everything (just look at Windows), it does not appear to be organic at all. It's not a cure all for everything. It may have it's place (won't be as glamorous as everyone is thinking) and we will settle to a couple of models (which has it's own problems).
Will it be here to stay.....yea, because corporations need it to stay after all of the money that they have poured into it.

Not all new things are actually good. It all depends on how things are implemented. Even if one has a good idea, how it's implemented can make it a crap one in a heartbeat.

Even with regard to the internet, while it has been a boon in some way, it sucks in others. Always sacrifices, the question one has to ask, is the juice worth the squeeze? With regard to what passes for "AI", I don't think that it is.

Deep down, I really think that's what we all want.

While there are those that want that, or they know that they can't do what they want without that type of hand holding, that is certainly not what I want. The journey, for me, has always been just as important as the destination. With people's lackluster attention span nowadays, most couldn't sit still to be able to learn subjects in the way our minds are supposed to learn. Not even getting into the reading problem that we have here stateside and how that's going to affect "prompting" as well.

I'm more of the traditional luddite. Not against technology in of itself, but how it is implemented, that's where the problem is (most people think luddites are against tech period, that really isn't the case, it's how it's implemented, that's the key, and the same reasons that were applied way back when, still hold water in today's world).

It's not quite there yet but in a short amount of time it's become quite impressive. We'll always point out the flaws.

It always seem to be quite impressive for the normies, but why are the people that are on the front lines of it, that it's required by their companies talking about efficiency losses and how they don't like working with it? Yes, people are lazy, especially devs, so again, why is it having to be forced on them if the tooling is just that good?

I can't wait until self driving cars are eventually required by law and by insurance companies and when deaths by automobile related accidents decrease by 98%, there will be that 2% that complain like it didn't actually "FIX" the problem. I hate driving in the world today. That part of AI can't come fast enough. (Sorry, that one was a rant)

I'm still waiting for flying cars that was promised to happen when I was just a little kid, still hasn't happened.

The problem that I have with vehicles as they are now, they are connected to the WAN (more to be a pain for right to repair in my estimation, but I'm speculating there), that is a horrible vector for issues, regardless of how good self driving is. Ripe for 3rd party exploitation and very very dangerous to everyone on the road (most(I actually think all) have been demos, but I do seem to recall a Jeep that actually ended up being an accident in a controlled setting). All the electronic tech on it as well, can actually cause issues on the road, something goes bad, a FICM goes bad, that's not good while driving down the road (be it you driving or your vehicle). With all the sensors that vehicles have, while when operating normally, they can provide a lot of help, when they are not (and it can happen at the drop of the hat, gotta love electrical gremlins) no bueno, again doesn't matter who or what is driving the vehicle. I have had a sensor almost cause a wreck 3 times, because it works within certain parameters (it didn't know context of the situation, which is key, no concept of context) and it wasn't the case. So no, I'm not to thrilled for that as there is no reason to think that it won't happen with "AI". Shoot in one area of "AI" on Windows, MS stated to be aware that it could download and run malware, imagine if that happened in a vehicle, unless you are assuming that everything is just going to be smooth sailing from the get go?

While I don't necessarily have a problem when someone doesn't like something, I certainly don't like having to deal with dipwads that can't drive like they have a brain cell when I'm in a long bed extended cab DRW with a 6 horse gooseneck trailer, however, I would not be clamoring for more regulation with regard to that. Seems I'm the one with the issue, I adjust accordingly. As the next person should. Very rarely does more regulation solve anything. I don't like driving in today's world either, but I'm not going to wish for more legislation, just because I don't like something and I want something to go towards my own comforts.

 

[binki]

So far chat has been really dumb with most of what I have asked it. Maybe it is just the way I ask. I have started asking multiple AI engines the same question to see if it is me or the engine. The best that it has done is to create a poster for me of a pinball machine back glass. It did pretty well with that. It has also helped me out modifying artwork faster than I could like removing words or replacing them. The only downside is it produces a png instead of vector art, but I will take it.

 

[CMYKENGINEERING]

I look forward to the future when all of us are useless because we can't figure out how to do anything without "AI".

 

[somcalmetim]

I asked ChatGPT for an outlook on a stock and I could tell the price it gave me was wrong.
After some prompting, it told me "Hey thanks for calling me out on that... sometimes to save processing power/time I just check a snapshot of the world I have from mid-2024 for stock and live crypto pricing" ...wtf??? That's not very f****** useful...
It told me to make sure that I told it that I wanted it to "use current data" in its advice...**facepalm**

 

[danhayes1188]

not good for much but ive used it to create extensive practice tests for journeyman and master sign electrician practice. i dont trust the answers but it gets me thinking regardless

 

[Stacey K]

Someone was asked the question (maybe it was to Elon Musk? I can't remember) "what happens when we are all replaced by robots". They replied back something like "that's the goal, the robots do all the work and everyone just lives without needing money" - or something like that, maybe everyone gets a base wage.

I'm a very surprised we aren't closer to that point.

One of my customers bought a robotic lawn mower that mows the soccer field and a couple other large properties. His goal would be to just have one guy and multiple robotic lawnmowers. So, I guess the technology is starting to show up in even basic jobs like lawncare. I imagine some of that stuff is so high priced it's tough for regular to buy, just like computers were back in the 80's and 90's.

When I was a kid, I loved The Jetsons, I thought for sure when I was old the world would look like that. Guess it takes longer to progress than my little mind thought LOL

 

[jwlllpl]

Still don't have that family of HoverCraft parked above my garage like the 80s promised me. Where did I go wrong?