Ai - The Final Frontier

[petrushkagoogol]

AlphaZero, the artificial intelligence program developed by Google, has obliterated the strongest chess engine Stockfish with 28 wins, 72 draws and no losses.

 

This marks a watershed moment in human history wrt to computers.

 

While some may choose to align themselves with the Prophets of Doom, I prefer to remain optimistic.

 

According to me AI is most useful where -

 

* there are many variables

* algorithms need to be dynamic

* algorithms need to be self-tweaking

 

Some good uses for this kind of computing could be -

 

* solving criminal cases

* econometric models using regression and correlation

* meterological forecasting to a high degree of accuracy.

 

Can you please share your thoughts on this ?

Edited December 8, 2017 by petrushkagoogol

[sanctus]

I like the saying about AI going something like: "AI excels in all the things we find diffcult and fails miserably in the things we think is easy".

Examples of where it excels are plenty: GO (=more combinations than atoms in universe), Chess, etc.
While recognising images (eg. horse or donkey on pic?) which is so simple that a 4 year old does it, is a a big challenge for AI.

Also you say "Some good uses for this kind of computing could be -" you should replace the "could be" with "are", it is already done.

[spartan45]

Google AI AlphaZero’s victory over the chess engine stockfish 8 is impressive. A balance to this is human computer programmer’s ingenuity. I downloaded the top rated stockfish 8 32bit chess engine and was amazed it is only 1,314KB, about the 1.44MB size of an old floppy disk! https://stockfishchess.org/download/

 

The ‘engine’ is put into a GUI (graphical user interface) or ‘car’ usually using the protocols WinBoard or UCI. The ‘car’ I used was Fritz 12 chess (runs Winboard protocol) which happily accepted the downloaded 1,314KB stockfish 8 32bit engine. Google ‘How much does the GUI affect the strength of a chess engine?’ to find the 2^nd top result: chess.com forum Lucas Chess. https://www.chess.com/forum/view/chess-equipment/lucas-chess-engine-settings-are-very-weak?page=3

 

This post is informative and ends at page 3 with this link http://lucaschess.pythonanywhere.com/

This GUI  ‘Lucas Chess’ is 55.2MB running UCI protocol and stockfish 8 with 39 other engines included! This is an ingenious chess program. Incidentally the GUI or ‘car’ only affects the strength of the engine if the GUI has default handicap settings not noticed by the operator, or are accidentally activated by said operator.

 

Human ingenuity shines ever bright.

[DingerM786]

What if you had the chance to create a AI, what would the AI's purpose be?

[Maine farmer]

What if you had the chance to create a AI, what would the AI's purpose be?

To supplant humanity, of course.    Intentional, or not.

[DingerM786]

To supplant humanity, of course.    Intentional, or not.

i agree, that is exactly what i would do but when i asked other people, they said they would make an AI that would supplant themselves in different ways, so what made you choose humanity over yourself? 

Edited December 20, 2017 by DingerM786

[DingerM786]

I like the saying about AI going something like: "AI excels in all the things we find diffcult and fails miserably in the things we think is easy".

 

Examples of where it excels are plenty: GO (=more combinations than atoms in universe), Chess, etc.

While recognising images (eg. horse or donkey on pic?) which is so simple that a 4 year old does it, is a a big challenge for AI.

 

Also you say "Some good uses for this kind of computing could be -" you should replace the "could be" with "are", it is already done.

What do you like about AI's

[Maine farmer]

i agree, that is exactly what i would do but when i asked other people, they said they would make an AI that would supplant themselves in different ways, so what made you choose humanity over yourself? 

It's not so much a choice, but it looks like a matter of inevitability.

 

I like the concept of mortality.  I've seen people at the end of their lives that were ready to go, and I'm sure I will one day reach that point.  

[DingerM786]

It's not so much a choice, but it looks like a matter of inevitability.

 

I like the concept of mortality.  I've seen people at the end of their lives that were ready to go, and I'm sure I will one day reach that point.  

right but if you had the choice, what would be, by the way i like the concept of morality too 

[Maine farmer]

right but if you had the choice, what would be, by the way i like the concept of morality too 

Well, I didn't mention morality but mortality.  Morals are another related topic...

 

As for choice, I like nature, but human nature may lead us inevitably towards the artificial.  A bit of a conundrum there.

Edited December 20, 2017 by Farming guy

[sanctus]

What do you like about AI's

Everything:-) I see the scientific advances for example or giving humans the possibility to do what they like or the fact of potentially becoming immortal or etc.

 

I do not share the fear some people have. A risk I see is more like the AIs in the movie HER, not like skynet.

 

 

[DingerM786]

Everything:-) I see the scientific advances for example or giving humans the possibility to do what they like or the fact of potentially becoming immortal or etc.

 

I do not share the fear some people have. A risk I see is more like the AIs in the movie HER, not like skynet.

 

 

if there was a chance to be immortal without the use of AI, would you attempt to be immortal knowing that it could put your life at stake?

[Maine farmer]

Everything:-) I see the scientific advances for example or giving humans the possibility to do what they like or the fact of potentially becoming immortal or etc.

 

I do not share the fear some people have. A risk I see is more like the AIs in the movie HER, not like skynet.

 

 

I'd be willing to bet that immortality would not be such a great thing.  I've noticed that as I age, it's easier for me to get "set in my ways" much as I try not to, and a fresh perspective  is more difficult for older folks.  After a while we would get bored.  You can only experience the first time once, after all.  

[sanctus]

But you could travel to other planets and come back and still know people :-) And since they are "set in their ways" you will still know them :D.

 

[DingerM786]

But you could travel to other planets and come back and still know people :-) And since they are "set in their ways" you will still know them :D.

 

 

I agree, you think that by traveling to other planets and coming back will still allow you and other people to recognise you