The AI Thread
- Anthony.R.Brown
- Veteran
- Posts: 157
- Joined: Thu Mar 27, 2014 1:03 pm
The AI Thread
This thread is all about AI (Artificial Intelligence) for those that don’t know
Below is a link explaining more about it…
en.wikipedia.org/wiki/Artificial_intelligence
And another view on it…
joshworth.com/stop-calling-in-artificial-intelligence
Related to this AI thread I am about to release the latest version of my OXAUTOLN3X3GENIUS Tic-Tac-Toe Auto learning program the perfect and final version
You might ask why the delay ? well I have had a week or two of inspiration and problem solving with it,where I solved three important things!,the last one turned out to be better left as it was ? Because it left out an important thing regarding learning,compared to how the Human brain works,even though we solve things we never delete the mistakes from our learning process ? And it’s how we can compare the past to the present and hopefully the future,and how we improve our learning.
The other reason is in this inventive creative two weeks ? I have invented a New type of AI Engine for my Tic-Tac-Toe program and other things! ,it will not replace the original as I am sure it’s not as good ? I will also try a version combining both the Engines,so I am busy until I have exhausted all possibilities.
Both will be in the program as a player option for Humans to play against,and for them to play each other as well as a Random player.
The name of my New AI Engine is WOPR(AI) (War Operation Plan Response) the Super Computer from the May 7,1983 film Wargames,where after it plays itself until near self destruction at Tic-Tac-Toe to a Draw! it makes the famous quote "The only winning move is not to play."
The Wargames link is below…
en.wikipedia.org/wiki/WarGames
And the famous WOPR Tic-Tac-Toe game is below…
www.youtube.com/watch?v=F7qOV8xonfY
Anthony.R.Brown
Below is a link explaining more about it…
en.wikipedia.org/wiki/Artificial_intelligence
And another view on it…
joshworth.com/stop-calling-in-artificial-intelligence
Related to this AI thread I am about to release the latest version of my OXAUTOLN3X3GENIUS Tic-Tac-Toe Auto learning program the perfect and final version
You might ask why the delay ? well I have had a week or two of inspiration and problem solving with it,where I solved three important things!,the last one turned out to be better left as it was ? Because it left out an important thing regarding learning,compared to how the Human brain works,even though we solve things we never delete the mistakes from our learning process ? And it’s how we can compare the past to the present and hopefully the future,and how we improve our learning.
The other reason is in this inventive creative two weeks ? I have invented a New type of AI Engine for my Tic-Tac-Toe program and other things! ,it will not replace the original as I am sure it’s not as good ? I will also try a version combining both the Engines,so I am busy until I have exhausted all possibilities.
Both will be in the program as a player option for Humans to play against,and for them to play each other as well as a Random player.
The name of my New AI Engine is WOPR(AI) (War Operation Plan Response) the Super Computer from the May 7,1983 film Wargames,where after it plays itself until near self destruction at Tic-Tac-Toe to a Draw! it makes the famous quote "The only winning move is not to play."
The Wargames link is below…
en.wikipedia.org/wiki/WarGames
And the famous WOPR Tic-Tac-Toe game is below…
www.youtube.com/watch?v=F7qOV8xonfY
Anthony.R.Brown
- Anthony.R.Brown
- Veteran
- Posts: 157
- Joined: Thu Mar 27, 2014 1:03 pm
Re: The AI Thread
Check out my latest OXAUTOLN3X3GENIUS Final version! below ...
A.R.B
A.R.B
Last edited by Anthony.R.Brown on Sun May 23, 2021 8:16 am, edited 1 time in total.
- Anthony.R.Brown
- Veteran
- Posts: 157
- Joined: Thu Mar 27, 2014 1:03 pm
Re: The AI Thread
At last! The Original OXAUTOLN1 23/02/2002 is now working 100% after removing the Arrays problems!
The New! Updated version of the program is attached OXAUTOLN1UPDATE,it is nothing fancy just basic play,first I had to get it to this stage,now I can remake the GENIUS version with all the bells & whistles,the main difference with the first GENIUS version,and the New one is that it will also be possible to play the Original against the GENIUS version
This version is now with all the other versions below...
A.R.B
The New! Updated version of the program is attached OXAUTOLN1UPDATE,it is nothing fancy just basic play,first I had to get it to this stage,now I can remake the GENIUS version with all the bells & whistles,the main difference with the first GENIUS version,and the New one is that it will also be possible to play the Original against the GENIUS version
This version is now with all the other versions below...
A.R.B
Last edited by Anthony.R.Brown on Mon May 24, 2021 2:07 pm, edited 1 time in total.
- Anthony.R.Brown
- Veteran
- Posts: 157
- Joined: Thu Mar 27, 2014 1:03 pm
- Anthony.R.Brown
- Veteran
- Posts: 157
- Joined: Thu Mar 27, 2014 1:03 pm
Re: The AI Thread
From my THE TIC-TAC-TOE THREAD below.. because it's AI related...
http://www.petesqbsite.com/phpBB3/viewt ... 104#p39104
Rowan Gilmore, studied at University of Cambridge
There are 255168 possible game of Tic-tac-toe excluding symmetry. The first player wins 131184 of these, the second player wins 77904 games and the remaining 46080 are drawn.
As has been pointed out, with best play all games should result in a draw. Hence although there are 209088 winning games, many of these would almost never occur in practice.
For those interested, the python code I used to simulate this is given below:
Would it be possible to convert the code below to run in QB64 ?
nWinO, nWinX, nDraw = 0, 0, 0
def recurse(board, toMove):
global nWinO, nWinX, nDraw
def win(board, player):
return (any(all(board[j] == player for j in range(3)) for i in range(3)) or
any(all(board[j] == player for i in range(3)) for j in range(3)) or
all(board == player for i in range(3)) or
all(board[2-i] == player for i in range(3)))
def draw(board): return all(board[j] != '' for i in range(3) for j in range(3))
if win(board, 'O'): nWinO += 1
elif win(board, 'X'): nWinX += 1
elif draw(board): nDraw += 1
else:
for i in range(3):
for j in range(3):
if board[j] == '':
board[j] = toMove
recurse(board, 'X' if toMove == 'O' else 'O')
board[j] = ''
recurse([['','',''],['','',''],['','','']], 'O')
print("There are %d possible games (excluding symmetry), of which O wins %d, X wins %d and %d are drawn." % (nWinO+nWinX+nDraw,nWinO,nWinX,nDraw))
http://www.petesqbsite.com/phpBB3/viewt ... 104#p39104
Rowan Gilmore, studied at University of Cambridge
There are 255168 possible game of Tic-tac-toe excluding symmetry. The first player wins 131184 of these, the second player wins 77904 games and the remaining 46080 are drawn.
As has been pointed out, with best play all games should result in a draw. Hence although there are 209088 winning games, many of these would almost never occur in practice.
For those interested, the python code I used to simulate this is given below:
Would it be possible to convert the code below to run in QB64 ?
nWinO, nWinX, nDraw = 0, 0, 0
def recurse(board, toMove):
global nWinO, nWinX, nDraw
def win(board, player):
return (any(all(board[j] == player for j in range(3)) for i in range(3)) or
any(all(board[j] == player for i in range(3)) for j in range(3)) or
all(board == player for i in range(3)) or
all(board[2-i] == player for i in range(3)))
def draw(board): return all(board[j] != '' for i in range(3) for j in range(3))
if win(board, 'O'): nWinO += 1
elif win(board, 'X'): nWinX += 1
elif draw(board): nDraw += 1
else:
for i in range(3):
for j in range(3):
if board[j] == '':
board[j] = toMove
recurse(board, 'X' if toMove == 'O' else 'O')
board[j] = ''
recurse([['','',''],['','',''],['','','']], 'O')
print("There are %d possible games (excluding symmetry), of which O wins %d, X wins %d and %d are drawn." % (nWinO+nWinX+nDraw,nWinO,nWinX,nDraw))
- Anthony.R.Brown
- Veteran
- Posts: 157
- Joined: Thu Mar 27, 2014 1:03 pm
Re: The AI Thread
Neural Networks: parameters, hyperparameters and optimization strategies
https://towardsdatascience.com/neural-n ... 0842fac0a5
A.R.B
https://towardsdatascience.com/neural-n ... 0842fac0a5
A.R.B
- Anthony.R.Brown
- Veteran
- Posts: 157
- Joined: Thu Mar 27, 2014 1:03 pm
The AI Thread
Below shows my THEARBTICTACTOEV3X3PROGRAM destroying another TIC-TAC-TOE program namely TIC TAC TOE by Paul Meyer & TheBOB...
Last edited by Anthony.R.Brown on Tue Jul 19, 2022 3:49 am, edited 2 times in total.
- Anthony.R.Brown
- Veteran
- Posts: 157
- Joined: Thu Mar 27, 2014 1:03 pm
- Anthony.R.Brown
- Veteran
- Posts: 157
- Joined: Thu Mar 27, 2014 1:03 pm
- Anthony.R.Brown
- Veteran
- Posts: 157
- Joined: Thu Mar 27, 2014 1:03 pm
- Anthony.R.Brown
- Veteran
- Posts: 157
- Joined: Thu Mar 27, 2014 1:03 pm
The AI Thread
Below attached is my THEARBTICTACTOEV3X3PROGRAM
A.R.B