Has Anyone Used ChatGPT For Help With Bonsai

Bonsai_hope

Yamadori
Messages
98
Reaction score
44
Location
Savannah, GA
USDA Zone
9a
I was fooling around with ChatGPT this afternoon and asked it for help with my Chinese Elm. It gave me very accurate advice. Then it asked me if I wanted suggestions wrt my particular tree. I uploaded a picture. This is what I got back. I’m flabbergasted.
 

Attachments

  • 5FA676F6-2D91-4FE9-B313-9B325C203002.png
    5FA676F6-2D91-4FE9-B313-9B325C203002.png
    1.6 MB · Views: 127
You can get really good responses from AI chatbots if you give it the right prompts. If you just ask something it will give you back whatever it finds first. It doesn't fact check, think about the context or check if the source is reliable. It could be correct or complete and utter nonsense. If you for example ask it to use info from specific sources or people you can improve a response tons.

I find it extremly useful and time saving, however the moral dilemmas are huge as all it does is gather info or content that people at some point have shared or created. The potential long term risks of AI are also substantial.

Also that picture makes no sense.
 
ChatGPT is GREAT for my job as a programmer. @Ply is right, though. You sometimes really have to chat back and forth with it to get what you need. You can tell it it's wrong or is giving you bad info, which will make it work a little harder.

I saw something interesting a while back... people were asking the AI to make a portrait of what it "thought" the user looked like. The portraits returned seemed to reflect the user's usage of it. Some of them were pretty scary, where the users obviously abuse the AI, only ask about horrific personal issues, or ask it to do weird things. 😳
 
You can tell it it's wrong or is giving you bad info, which will make it work a little harder.
Only problem is if you're not an expert at something (which is why you'd use it), you wouldn't know if its right or wrong, and if AI chatbots are good at one thing its being confident in what they're saying whether its correct or not. Ditto to AI tools like copilot saving time with programming though.

I uploaded a picture. This is what I got back.

The image it spits out is similar to the the type of images that people like here put up (as in arrows, and defining rounded lines), but it doesn't really mean anything. For your example, what are the arrow pointing at? The dotted line doesn't really make sense either. Seems like a weird place to define a pad or canopy, and that's certainly not where you'd want to prune if you're trying to ramify those branches. ChatGPT may be convenient, but why use it when you can come to a site like this and ask some of the actual people who unwillingly contribute to the chatbots?
 
You can get really good responses from AI chatbots if you give it the right prompts. If you just ask something it will give you back whatever it finds first. It doesn't fact check, think about the context or check if the source is reliable. It could be correct or complete and utter nonsense. If you for example ask it to use info from specific sources or people you can improve a response tons.

I find it extremly useful and time saving, however the moral dilemmas are huge as all it does is gather info or content that people at some point have shared or created. The potential long term risks of AI are also substantial.

Also that picture makes no sense.
I’m curious, what in that picture makes no sense?
 
I don't use those things, they are just garbage. I would also encourage you to stop using it. Outsourcing your thinking is generally a bad idea.
Funny story...

I rented a little cabin in the woods a few weeks back. Imagine my surprise when, on the day I was going to drive to the cabin, I found that three out of the four roads leading to the cabin were closed due to rain damage (from a hurricane that occurred over a year ago). There was no mention of access issues on the listing, and I would have had to drive 24 miles out of my way (on mountain roads) in order to get to the cabin, and the same access issues any time I wanted to leave to go to a restaurant or go shopping.

When I complained to the landlord, she was very cold, and explained that "according to ChatGPT" it was "only going to be ten minutes out of my way".

Tell me the mode of transportation that can drive 24 miles on winding mountain roads with a 45 mph speed limit in 10 minutes... :) Perhaps ChatGPT was referring to helicopter? Who knows? But the irony is that someone didn't question the response - and was quoting it as fact. The single greatest concern I have with "AI" is that the people most likely to refer to it are those least likely to understand the response they receive.
 
Last edited:
Not comepltee nonsense tbh:

I wonder what BonsaiEmpire thinks about your chatbot stealing and reposting all of their intellectual property? I certainly don't remember it asking me for permission to take content from this site and republish it as its own - while selling ads and earning revenue by doing so.

Note also the comment that it says is coming from BonsaiNut. "For tropicals: some growers use lava + turface + pine bark + akadama + grit" Oh rly? This is wrong... and I'm not willing to post why and give all of that knowledge to a chatbot - but it is something that most beginners would even say "this doesn't make sense". Not too bad if you are talking bonsai. Really bad if you are using AI to drive a car, or suggest medical care.

Separately, your chatbot's "intelligence" stops the moment it can't steal any more information. ChatBot "AI" is based solely on history - and its ability to steal human thoughts that have been published somewhere. A chatbot cannot tell you what I am thinking, or for that matter, what the doctor about to perform surgery on you is thinking. It cannot reason. It cannot explore options, test hypotheses, adjust based on reality. It can only regurgitate.
 
The single greatest concern I have with "AI" is that the people most likely to refer to it are those least likely to understand the response they receive.
Outsourcing your thinking is generally a bad idea.
Spitting some real wisdom here! I wonder what this implies when applied to all the "professionals" who use it for "work".
 
I had ChatGPT search for a specific size range of pumice with cost analysis of the product including shipping.

I could have done this myself albeit in a much longer time frame.

I generally use it to search for a generic question that I don't know how to ask with search engines or to do a massive research task for me for a better spring board into the topic.

It seems very useful, but I would always be cautious of what it responds with. AI is incredibly intelligent, but it is not very wise in their current iteration.
 
Spitting some real wisdom here! I wonder what this implies when applied to all the "professionals" who use it for "work".
I did use it once to help explain a nuance concept to a coworker about my machining trade. To give a brief and vague idea, my experience position a cutter a specific way but I could not explain why various conflicting ideas/practices in the trade produced the desired result.
 
Back
Top Bottom