Just intrigued to hear people’s experiences, since AI seems to be pervading most aspects of our life in one way or another. I’ve never tried but wondering if chatgpt is able to look at grow pics and offer advice and that kind of thing, or make a weekly schedule, or troubleshoot, or whether instead this is one of the areas its totally useless / slightly undesirable in …
I would not use AI for a specific ANYTHING let alone cannabis.
What " AI " is good for is to get you bulk information that you can use to study and become familure with growing pot.
Do not ask " how do I grow pot."
Ask it " post some links on well respected cannabis grow sites that explains the difference on how to water soil with organics and coco with mineral. "
There are a few people I know that have used and actually posted some of the replies from the program. It is or seemed to me to be more focused on giving the user a good experience with lots of " atta boy " and my favorite " perfect grower intuition" comments that really got creepy quickly.
Power tip: You can ask some of the programs to stop any kind of interpersonal comments and deliver in a dry factual tone that reminds me a lot of how I imagine Spock would answer if he was helping a peep grow pot.
I have heard some horror stories of grows gone terribly wrong as a result of relying on AI’s advice
it’s definitely worth taking the slightly longer route of getting guidance and advice from actual growers with real life stories. @Pothead64 's suggestion of asking it to source respected sites is golden ![]()
Great response and really helpful advice. Basically how I use it in my own life as well … got a little bit bored of how sycophantic it is for lots of things and can see it being a similar thing with growing weed…
Lol - pains me to think more generally of all the awful decisions people have made so far in life that have been as a result of chatbots ![]()
Helpful for sure in many instances but god awful in others
One of my last and best jobs ever was at a restoration and modification shop where I had lovely tasks like " find this left side Chrysler hood emblem that is longer than the stock photo one readily available from NOS stock or side molding for a 63 Chevy or just whatever.
It took me years to develop the links and contacts for weird stuff.
Now with a detailed ask I get the same phone numbers and addresses it took all those years.
Not fu••in fair![]()
Strange times…
The large language models are not good at deciding. The regurgitate info already available and have no discernment of good or bad data. I have gotten wrong answers, told the LLM it was wrong and then it agrees with me, but I have to tell its error.
And it no doubt used language like " you are right to call me out on that error " and actually make the same one 3 days later.
Yeah I test this out sometimes too. There’s a guy on Instagram somewhere who talks to his AI and exposes how awful (and surprisingly stupid) it can be. It’s hilarious
Dont and wont use it ![]()
![]()
its right up there with turkey bacon![]()
![]()
I’ll do the latter ![]()