Log In Register

THE TRAINWRECK. Brought to you by thcfarrmer…..

My problem with ai is when its right its right but when its wrong doesntvsay i dont know but makes shit up.and now you trust it for years and then it fucks you one day. Similarry thats one of the things i hate in people as well when they cant say i dont...
Home Forums Medical Cannabis Cultivation Grow Diaries THE TRAINWRECK. Brought to you by thcfarrmer…..
Grow diary eligible · Medical Cannabis Cultivation

THE TRAINWRECK. Brought to you by thcfarrmer…..

by Captspaulding · Started
1d
Running
0
Updates
95,543
Replies
0
Images
Discussion below · 95,543 replies
Page 3953 of 4778 · Replies 79,041–79,060 of 95,544
Status
Not open for further replies.
My problem with ai is when its right its right but when its wrong doesntvsay i dont know but makes shit up.and now you trust it for years and then it fucks you one day. Similarry thats one of the things i hate in people as well when they cant say i dont know.
Have you seen the explosion of weenie clauses in "AI "answers? Now even in the text of the answers you see wishy-washy bullshit like " no common consensus" or " "further study may be needed *
Menu driven diagnostics have been tried for years as a well trained technician is very expensive. GM tried this by backing down on tech training and increasing menu driven when they launched the new scan tool that replaced the tech2.
Remember AI is nothing more than a string of " if B=$thengoto239804 " with a lot of memory and speed
Now the fun part. A recent review of 2900 masters degrees final thesis found that 20% of them were claptrap bullshit with huge foundational inaccuracies of basic principles of STEM.
And they are now integrated into AI.
From 2002 through 2007 an internal audit of GM found 15% of all technical information specs were not accurate and had to be " updated "
Every last bit of information released by the " government " has been washed in the machine. More invalid points.
AI will fail due to inputs. Incorrect data.
 
You should read what silica does

My problem with ai is when its right its right but when its wrong doesntvsay i dont know but makes shit up.and now you trust it for years and then it fucks you one day. Similarry thats one of the things i hate in people as well when they cant say i dont know.
When absolute correctness needed I use the prompt "I want your reply to have a 99.98% assurance of being factual and correct. If not 99.98% sure; scour the web and or ask me questions until the threshold is met" then ask the question. I can't say it's 100%... but it's usually pretty close.
 
1756292762081
 
Similarry thats one of the things i hate in people as well when they cant say i dont know.
Not only do I also hate this I just do not understand it.
Every thing I learned about vehicle repair I learned from someone else. From the guy who showed me how 5 minutes with a die grinder notching a front engine mount to save 2.4 hours on a 5 hour water pump to the last class I had on Telemetric security issues for Gated Buss systems information is always changing. Even a student with an outstanding education can learn wrong information.
 
Even a student with an outstanding education can learn wrong information.
Precisely why looking at building a rolling plant cage starts with "I have used a hammer before and I know what inches are", otherwise you won't know AI just told you to build a fucking brinks level security cage when you don't need it. The uneducated dingus will just build the time-waster.
 
Not only do I also hate this I just do not understand it.
Every thing I learned about vehicle repair I learned from someone else. From the guy who showed me how 5 minutes with a die grinder notching a front engine mount to save 2.4 hours on a 5 hour water pump to the last class I had on Telemetric security issues for Gated Buss systems information is always changing. Even a student with an outstanding education can learn wrong information.
No thats okej.but there exist people that think just because you asked you dont know and anything they say its gonna be good. Hard to explain it but i dont mean if i tell ?ou wrong and you krep spreading it.its they know you dont know and just lie to be smart. Maybe you never met a person that his whole life never said well i dont know.
 
No thats okej.but there exist people that think just because you asked you dont know and anything they say its gonna be good. Hard to explain it but i dont mean if i tell ?ou wrong and you krep spreading it.its they know you dont know and just lie to be smart. Maybe you never met a person that his whole life never said well i dont know.
Hehe.
I have made teaches and instructors say " I don't know " many times.
The key is if they follow the declaration with " but I will find out for you".
 
Have you seen the explosion of weenie clauses in "AI "answers? Now even in the text of the answers you see wishy-washy bullshit like " no common consensus" or " "further study may be needed *
Menu driven diagnostics have been tried for years as a well trained technician is very expensive. GM tried this by backing down on tech training and increasing menu driven when they launched the new scan tool that replaced the tech2.
Remember AI is nothing more than a string of " if B=$thengoto239804 " with a lot of memory and speed
Now the fun part. A recent review of 2900 masters degrees final thesis found that 20% of them were claptrap bullshit with huge foundational inaccuracies of basic principles of STEM.
And they are now integrated into AI.
From 2002 through 2007 an internal audit of GM found 15% of all technical information specs were not accurate and had to be " updated "
Every last bit of information released by the " government " has been washed in the machine. More invalid points.
AI will fail due to inputs. Incorrect data.
Sorry bro, but I have to respectfully disagree. AI has been slowly and quietly evolving for over 30yrs. We just used to call it algorithms. Now it's growth and capabilities are very near vertical on an upward exponential curve. Not only will it not fail, but when ASI gets here, we're either going to be saved or doomed depending on who achieves it 1st, and what it's disposition turns out to be. Whomever comes in 2nd will never ever, ever be able to catch up.

There is no doubt that it will continue to have hiccups and flaws the same as any new tech, or systems, but the overall intelligence trend will continue ever upward at an ever increasing rate. Pandora is out of the box...

Mankind has been saying "We'll never achive this, or create that..." since we learned to communicate. Near every instance has proven false. GAI 1st, then ASI will be here in no time... When ASI gets here, worldwide doubling time of information will be so quick, we don't even have time concepts to describe it.

Just because you don't understand it or can cite insurances of failure, doesn't doom it to failure, in any way. Hell, the creators don't even know how it works now...

crazy times to be alive... ✌️
 
Hehe.
I have made teaches and instructors say " I don't know " many times.
The key is if they follow the declaration with " but I will find out for you".
My fav teacher taught us that questions are more important than answers... 🤓👍
 
Sorry bro, but I have to respectfully disagree. AI has been slowly and quietly evolving for over 30yrs. We just used to call it algorithms. Now it's growth and capabilities are very near vertical on an upward exponential curve. Not only will it not fail, but when ASI gets here, we're either going to be saved or doomed depending on who achieves it 1st, and what it's disposition turns out to be. Whomever comes in 2nd will never ever, ever be able to catch up.

There is no doubt that it will continue to have hiccups and flaws the same as any new tech, or systems, but the overall intelligence trend will continue ever upward at an ever increasing rate. Pandora is out of the box...

Mankind has been saying "We'll never achive this, or create that..." since we learned to communicate. Near every instance has proven false. GAI 1st, then ASI will be here in no time... When ASI gets here, worldwide doubling time of information will be so quick, we don't even have time concepts to describe it.

Just because you don't understand it or can cite insurances of failure, doesn't doom it to failure, in any way. Hell, the creators don't even know how it works now...

crazy times to be alive... ✌️
AI will make people more stupid. AI will do the same thing as cell phones have done which is destroy memory functions by not using them any more.
AI can and will and has been corrupted by data input. Versions have been built with clear political bias and the answers show this.

Dude totally not a Luddite against tech but this has all the ingredients of a massive mistake.
 
Does everyone know what time it is?
View attachment 2509261
Now a question. Have not started smoking yet.
Is it even legal to make special brownies without being high?
Done.
These are extra spicy. My normal add is 7 grams and it was to little for one brownie but to much for 2.
This has 10 grams apple fritter coffee ground and decarbed.
IMG 20250827 080339819 HDR
 
When you find a secret page...

I needed it, fo reelz:

3 day shipping.
 
Last edited:
AI will make people more stupid. AI will do the same thing as cell phones have done which is destroy memory functions by not using them any more.
AI can and will and has been corrupted by data input. Versions have been built with clear political bias and the answers show this.

Dude totally not a Luddite against tech but this has all the ingredients of a massive mistake.
Oh it's definitely a mistake in the short term, long term remains to be seen. That's mostly what I've been trying to convey. As I've already said: the biggest problem I see are bad actors doing really bad things with it
 
Status
Not open for further replies.
Page 3953 of 4778 · Replies 79,041–79,060 of 95,544
Back
Top Bottom