Live Stoner Chat Live Stoner Chat - Apr-Jun '24

[WildBill]

Woke up and it's a bit chilly in the house! I guess that's OK though. We're going to have about a 40*F temperature rise today!
Next four or five days temperatures will be hovering around 80s at a high. After a slight dip in highs for three days, with highs hovering around 80 and nights in the 60s, I could easily harden off some 4 to 6 week old tomato plants or cannabis and get a nice little jump.

Well I didn't get any 'maters planted and I wish I could Legally grow outside and not worry!

Going to be some nice open window weather!


And best of all, no energy costs!



That's the Inner @Fermented_Fruitz coming out!

 

[WildBill]

OK here's some rabbit hole material!

" Grafting cannabis................. Can it be done and what can be done?"

I have played around with grafting on fruit trees and pecans. It's just something that popped in my head.

 

[Mossy]

Well I didn't get any 'maters planted

Nor did I......... ... but.... I have mine trained to sort themselves out....

See them in amongst the self germing nasturtiums......?........

They prefer me not to touch them.......... just add watter..........,

 

[Fermented_Fruitz]

Being baked is the only thing that helps me deal with the public

 

[Fermented_Fruitz]

Good Morning everyone

 

[Olderfart]

Interesting read:

AutoBNN: Probabilistic time series forecasting with compositional bayesian neural networks

blog.research.google

I've done Bayesian modeling, as well as other types of computer modeling, and think I understand its limitations. "Garbage in - garbage out" comes to mind. The end results depend sensitively on the input, and in the specific case of Bayesian approaches, the input is often educated guesswork of general relationships between variables, not strong data about them. The tool can be used with care for exploring sensitivity of modeled systems to particular influences, but pretending that accurate future predictions can be generated is in my opinion optimistic. The modified approach described in the paper seems to intend to improve the Bayesian process by using observations to adjust the Bayesian functions, but predictions based on Bayesian models will still be limited by the underlying weakness of applicable data. The very use of the Bayesian approach is admission that data which would allow more precise modeling approaches are not possible due to lack of good data. They too often in my opinion dress up somewhat educated guesses as something more reliable. Basing important decisions on them is risky.

Bayesian modeling discussions on a stoner forum. Who woulda thunk?

 

[Olderfart]

@VitaMan This morning when I sat down to look at the news, I took two big dab hits and came across that article. I am reading it and think I am following along and all of the sudden I am down the rabbit hole, I did not have a clue WTF they were talking about

And here I thought you were serious... Stoners eh?

 

[Fermented_Fruitz]

View attachment 1675245
Just a thought, but perhaps you should hang this up…

 

[Fermented_Fruitz]

Knocking out some sets of pushups before heading out for our walk

 

[Fermented_Fruitz]

" Grafting cannabis................. Can it be done and what can be done?"