Has anyone ever grown with Ai

It’s more an Automated Intelligence than an Artificial Intelligence…. You’re not learning, you’re just mimicking a method based on statistics. You won’t learn to develop your instincts and feelings your plants… instantly telling you something is wrong by the stress tension that you can feel when you make the first step in the room when light turns on … NO AI can learn you this.

I use just I.

I've learned a lot through AI as well, but the more you read outside of AI, talk to other growers, and watch grow videos, the more you realize you shouldn't trust AI too much. It can give incorrect information, especially if you don't ask very specific questions. For example, when you're using organic nutrients like BioBizz, AI will sometimes give advice that applies to mineral feeding, organic feeding, or a mix of both. You really need to be precise and mention multiple times what nutrients and growing method you're using. AI is a great learning tool, but I always double-check important information with experienced growers and other sources.

Haha, accurate! No, I’ve mostly learned from my mistakes 🤣🤣🤣"

There's a good YT channel called Weed in a Pot (and sister channel called Hemp in a pot) and he did a complete grow a year or two ago using AI.

in what sense are you meaning "AI"? using a bounded LLM with only primary source material you give it like now defunct versions of notebook lm? heckin yep using a bespoke imaging model like open source and not-black box user-trained imaging models for tissue analysis or spore discernment? heckin yep using machine learning algorithms on your gear to converge on better settings, routines, schedules, thresholds, set points, etc. ? heckin yep copy someone else's question into an AI and paste the answer here? that is neon afro and unicycle level of clownage when most people think of AI they are going to be referring to LLMs of the mid 2020s...... this tek isnt new. google developed it many years ago and the papers on how to do it have always been free, because google did not see any use in something without any veracity sentence over sentence other than NPCs in games. ill be the first to admit that there an incredible amount of raw power in LLMs, but ill come back to that. LLMs (despite all the fancy user interface and buzzwords and venture capital circle jerks....) are nothing more than a multi-dimensional graph of a large order where the basic linguistics are bastardized mathematically. What it spits out is simply the output from the starting point on the graph you give it. you can get really specific prompting, you can train really really really fucking well, you can add giga watts of energy to compute layers for context.....but it is a graph bro. even if you cant visualize your prompt being plotted on a 15th dimensional graph and the resulting output based on statistics, linguistics, and training data combined...doesnt mean it isnt a graph. it still is. so when 9/10 people around you refer to AI, they are referring to a mid-grade graph and the major models available to the public are NOT trained well. they are reddit users and us shooting the shit on our alt accounts being gobbled up for training data lmao. the major models available to the public are NOT giving you 1000 times more compute just on context alignment. the raw power of LLMs are twofold. firstly, being able to replicate the nueron and synapse form of a brain is incredibly accurately done by the weighting and such of LLMs. second, linguistics is finite. the same way you could decipher some degree of communication between two verbal; species with different languages, so to will linguistics show certain phenoms once the LLMs are sufficiently large and have proper training data. that just isnt right now. not for you, not on chat gpt freemium or whatever bro. just imagine the wild degree of answers you get to a grow question on this site. now imagine each of those answers given equal weight and getting pumped into your statement. now, how many of those responses are terribly wrong full stop? lol

Hi.... Ich denke das man sich ein paar nützliche Tips holen kann, aber verlassen würde mich nicht drauf....... Denke die Tips hier bei Gd oder anderswo von realen Usern sind viel wertvoller ! Am besten sind natürlich die Erfahrungen, die du selbst sammelst und dann auch umsetzt.......😎 Viel Glück

Fuck AI!

I professionally use various models and while it improved a lot over the past 1-2 years, it still can't be relied on as a single source of truth. These systems still can and often generate very bad advice even about simple things.

The problem is it's trained on anecdote. So, don't assume it always has the right information.. It'll repeat popular sentiment if it is trained on user-submitted data - like forums and such. The LLMs are also not "AI." They are coded to be positive and make you feel good. So, there's that sort of corruption to the data, too. It will tell you what you want to hear if you cue it to do so.