Random Thought Thread

Thank you, CPK folk, for your recommendation of Sagewood for sheath danglers. It appears to do what I want. For the benefit of any who search in the future, the Sagewood 'medium' sheath adapter works on the current mashed cat FK3/BFK sheath in the same holes as the factory belt loop.
 
I know that it is fun and fashionable to beat up on AI and gawd only knows that I do plenty of that but that is because I am so underprepared and undereducated about this transformative technology. Truth is that in order to communicate efficiently or proficiently with any entity, be it anything of sentient nature or any machine, one has to know the language which that being understands or is programmed to understand. One can pretty much communicate with a chicken in the same way that one can communicate with a hammer, i.e, about pretty much nothing other than being of use and purpose.

Case in point, just like you can not be dropped in a foreign non English speaking country expecting to communicate with its people while not knowing much about their native tongue, you also can not be put in front of one of Nathan's computers while not knowing CAD to be expected to come up with a program to design a knife! AI uses complex mathematics to learn our language and as it is getting more and more proficient, the token technology which it uses to make its guesses statically more accurate it will require you to understand more about how to communicate with it more efficiently to harness the expanding power of those tokens.

I don't think that it is very efficient to use an AI platforms like ChatGPT or Gemini in the same manner as typing something very lazy and pedestrian in a Google search.

#ainathan
 
Casinostocks said:
I know that it is fun and fashionable to beat up on AI and gawd only knows that I do plenty of that but that is because I am so underprepared and undereducated about this transformative technology. Truth is that in order to communicate efficiently or proficiently with any entity, be it anything of sentient nature or any machine, one has to know the language which that being understands or is programmed to understand. One can pretty much communicate with a chicken in the same way that one can communicate with a hammer, i.e, about pretty much nothing other than being of use and purpose.

Case in point, just like you can not be dropped in a foreign non English speaking country expecting to communicate with its people while not knowing much about their native tongue, you also can not be put in front of one of Nathan's computers while not knowing CAD to be expected to come up with a program to design a knife! AI uses complex mathematics to learn our language and as it is getting more and more proficient, the token technology which it uses to make its guesses statically more accurate it will require you to understand more about how to communicate with it more efficiently to harness the expanding power of those tokens.

I don't think that it is very efficient to use an AI platforms like ChatGPT or Gemini in the same manner as typing something very lazy and pedestrian in a Google search.

#ainathan
Click to expand...
Very good read!

(There's a lot of words there so I skimmed through)

What I've learned by my quick read through..... AI can communicate with a hammer and a chicken

Thank you!
 
yoko said:
Very good read!

(There's a lot of words there so I skimmed through)

What I've learned by my quick read through..... AI can communicate with a hammer and a chicken

Thank you!
Click to expand...

What did Matt Gregory say about people not having the tendency to read any more just a few days ago? Oh! NVM!

You got the gist of my post wrong though! In essence, it meant that AI will be able to cook the best chicken dinner using a hammer if only you knew how to communicate with it proficiently ;)
 
Casinostocks said:
What did Matt Gregory say about people not having the tendency to read any more just a few days ago? Oh! NVM!

You got the gist of my post wrong though! In essence, it meant that AI will be able to cook the best chicken dinner using a hammer if only you knew how to communicate with it proficiently ;)
Click to expand...
I'm glad you didn't say " AI would start hammering chickens" Hahahahahaha
 
Casinostocks said:
I know that it is fun and fashionable to beat up on AI and gawd only knows that I do plenty of that but that is because I am so underprepared and undereducated about this transformative technology. Truth is that in order to communicate efficiently or proficiently with any entity, be it anything of sentient nature or any machine, one has to know the language which that being understands or is programmed to understand. One can pretty much communicate with a chicken in the same way that one can communicate with a hammer, i.e, about pretty much nothing other than being of use and purpose.

Case in point, just like you can not be dropped in a foreign non English speaking country expecting to communicate with its people while not knowing much about their native tongue, you also can not be put in front of one of Nathan's computers while not knowing CAD to be expected to come up with a program to design a knife! AI uses complex mathematics to learn our language and as it is getting more and more proficient, the token technology which it uses to make its guesses statically more accurate it will require you to understand more about how to communicate with it more efficiently to harness the expanding power of those tokens.

I don't think that it is very efficient to use an AI platforms like ChatGPT or Gemini in the same manner as typing something very lazy and pedestrian in a Google search.

#ainathan
Click to expand...

The problem is not the LLMs and the underlying computer science. The problem, IMO is that people trust them blindly, as they are used in platforms like google, etc.; even though they suck up unreviewed information and spit it back at you without verification and the problem of hallucinations has not been solved yet. Add to that:

- biased information, for example from Wikipedia; references are often untraceable, or worse faked by the models.
- scientific peer reviews are done largely with LLMs now, enforcing the existing science status quo and helping to ridicule new ideas.
- there is an obvious IP problem. For example, in my article above, on the left side a true cook, on the right side, an AI cook. Did AI violate the picture copyright ?

1764188216716.png

Even if the user is diligent and tries to verify, e.g., ChatGPT returns, they are sometimes near impossible to check, so I rather not use them in the first place.

Etc. I'll step off my box now.

Roland.
 
ferider said:
The problem is not the LLMs and the underlying computer science. The problem, IMO is that people trust them blindly, as they are used in platforms like google, etc.; even though they suck up unreviewed information and spit it back at you without verification and the problem of hallucinations has not been solved yet. Add to that:

- biased information, for example from Wikipedia; references are often untraceable, or worse faked by the models.
- scientific peer reviews are done largely with LLMs now, enforcing the existing science status quo and helping to ridicule new ideas.
- there is an obvious IP problem. For example, in my article above, on the left side a true cook, on the right side, an AI cook. Did AI violate the picture copyright ?

View attachment 3036013

Even if the user is diligent and tries to verify, e.g., ChatGPT returns, they are sometimes near impossible to check, so I rather not use them in the first place.

Etc. I'll step off my box now.

Roland.
Click to expand...
Her husband Bjork will have something to say about this
 
Last edited:
You must log in or register to reply here.
Back
Top