Uncategorized, chess, artificial intelligence, games, puzzles

Computer-Generated Chess Problem 02439

A newly published and original KBBNNP vs krbp chess problem generated by the prototype computer program, Chesthetica, using the Digital Synaptic Neural Substrate (DSNS) computational creativity approach. It doesn’t use endgame tablebases, deep learning or any kind of traditional AI. Chesthetica is able to generate mates in 3, mates in 4, mates in 5, study-like constructs and also compose problems using specific combinations of pieces fed into it (e.g. to compose something using a rook, bishop, knight and three pawns vs. a queen and a rook). Read more about it on ChessBase. Any chess position over 7 pieces could not possibly have been derived from an endgame tablebase which today is limited to 7 pieces.

kr6/3B4/bP2N3/1p6/8/3N1K2/1B6/8 w – – 0 1
White to Play and Mate in 5
Chesthetica v10.82 : Selangor, Malaysia
2018.12.18 5:23:13 AM
Solvability Estimate = Moderate

Chess puzzles are ancient. Some are over a thousand years old but only in the 21st century have computers been able to compose original ones on their own like humans can. Okay, let me think for a minute if there’s anything else to say here. Do share and try out some of the others too. Take some time to study the analysis and you might appreciate the puzzle a little more.

A Similar Problem by Chesthetica: 02034

Main Line of the Solution (Skip to 0:35)

Facebook | Instagram | Twitter | Book | Website

artificial intelligence, chess, games, puzzles, Uncategorized

Computer-Generated Chess Problem 02438

Contemplate this ‘KQRBPP vs kbbnpppp’ mate in 5 chess puzzle created by the program, Chesthetica, using the ‘Digital Synaptic Neural Substrate’ computational creativity method. It does not use endgame tablebases, artificial neural networks, machine learning or any kind of typical AI. The position below contains 14 pieces which means it simply could not have been derived even from an existing endgame tablebase which is presently limited to 7 pieces.

b4b2/3p1B2/nQ2R3/3pP3/k7/p1Pp4/7K/8 w – – 0 1
White to Play and Mate in 5
Chesthetica v10.82 : Selangor, Malaysia
2018.12.17 7:52:47 PM
Solvability Estimate = Difficult

If you notice any version of Chesthetica ‘skipped’ from one problem to the next, that simply means additional (minor) changes were made to the program before it was set to run again. White is over a rook’s worth in material but the precise win in this position still needs to be found. Why not time yourself how long it took you to solve this? Take some time to study the analysis and you might appreciate the puzzle a little more.

Main Line of the Solution (Skip to 0:35)

Amazon | BitChute | Minds | Reddit | Steemit | YouTube

artificial intelligence, chess, games, puzzles, Uncategorized

Computer-Generated Chess Problem 02436

A new if not unique KBNN vs kbn chess puzzle created by Chesthetica using the Digital Synaptic Neural Substrate AI computational creativity method. Chesthetica can compose problems that might otherwise take centuries or longer for human composers to think of, so you may enjoy them right now. Depending on the type and complexity of the problem desired, a single instance of Chesthetica running on a desktop computer can probably generate anywhere between one and ten problems per hour.

8/4B2b/8/8/N7/8/n2KN3/1k6 w – – 0 1
White to Play and Mate in 5
Chesthetica v10.86 : Selangor, Malaysia
2018.12.17 2:20:48 PM
Solvability Estimate = Difficult

Humans have been composing original chess problems for over a thousand years. Now a computer can do it too. Get a glimpse into the ‘mind’ of a computer composer. Did you find this one interesting or have something else to say? Leave a comment below! Collectively, these puzzles are intended to cater to players of all levels.

Main Line of the Solution (Skip to 0:35)

Amazon | BitChute | Minds | Reddit | Steemit | YouTube

artificial intelligence, chess, games, puzzles, Uncategorized

Computer-Generated Chess Problem 02432

Now, here we have a ‘KRBBP vs knp’ chess problem generated autonomously by Chesthetica using the Digital Synaptic Neural Substrate (DSNS) computational creativity approach. It doesn’t use endgame tablebases, neural networks or any kind of machine learning found in traditional AI. You can learn more about the DSNS here. The largest endgame tablebase in existence today is for 7 pieces (Lomonosov) which contains over 500 trillion positions, most of which have not been seen by human eyes. This problem with 8 pieces goes even beyond that and was therefore composed without any such help.

6B1/B3k3/4P1K1/6p1/2n5/8/8/1R6 w – – 0 1
White to Play and Mate in 5
Chesthetica v10.82 : Selangor, Malaysia
2018.12.15 2:13:55 AM
Solvability Estimate = Difficult

Get a glimpse into the ‘mind’ of a computer composer. Leave a comment below, if you like. Take some time to study the analysis and you might appreciate the puzzle a little more. If you’re bored of standard chess, though, why not try this?

A Similar Problem by Chesthetica: 00486

Main Line of the Solution (Skip to 0:35)

Amazon | BitChute | Minds | Reddit | Steemit | YouTube

artificial intelligence, chess, games, puzzles, Uncategorized

Computer-Generated Chess Problem 02431

This is an original ‘KQBBN vs kqnppp’ #5 chess problem generated autonomously by a computer using the ‘DSNS’ computational creativity approach which does not use any kind of machine or deep learning. Chesthetica is able to generate mates in 3, mates in 4, mates in 5, study-like constructs and also compose problems using specific combinations of pieces fed into it (e.g. to compose something using a rook, bishop, knight and three pawns vs. a queen and a rook). Read more about it on ChessBase. The largest endgame tablebase in existence today is for 7 pieces (Lomonosov) which contains over 500 trillion positions, most of which have not been seen by human eyes. This problem with 11 pieces goes even beyond that and was therefore composed without any such help.

8/1NK4Q/8/3k3B/3p2n1/4q1B1/2pp4/8 w – – 0 1
White to Play and Mate in 5
Chesthetica v10.82 : Selangor, Malaysia
2018.12.14 8:43:02 PM
Solvability Estimate = Easy

Some of the earliest chess problems by humans are over 10 centuries old but original ones by computer are very recent. Okay, let me think for a minute if there’s anything else to say here. Do share and try out some of the others too. Feel free to copy the position into a chess engine and discover even more variations of the solution.

A Similar Problem by Chesthetica: 01350

Main Line of the Solution (Skip to 0:35)

Facebook | Instagram | Twitter | Book | Website

artificial intelligence, chess, games, puzzles, Uncategorized

Computer-Generated Chess Problem 02425

Take a look at this ‘KBBBNNP vs kppp’ five-move chess puzzle or problem (whichever you wish to call it) composed by the prototype computer program, Chesthetica, using the approach known as the DSNS from the sub-field of AI, computational creativity. There is no known limit to the quantity or type of compositions that can be generated.

8/2B5/2p5/1B2P3/6N1/p6k/p1B1K3/3N4 w – – 0 1
White to Play and Mate in 5
Chesthetica v10.82 : Selangor, Malaysia
2018.12.9 10:50:44 PM
Solvability Estimate = Difficult

Chesthetica composes only unique or new constructs. If you have seen it before, cite the source and comment below because it is purely coincidental. Try to solve this puzzle. Do try some of the others in the series as well before you go. As a whole, these problems are intended to cater to players of all skill levels.

Main Line of the Solution (Skip to 0:35)

Amazon | BitChute | Minds | Reddit | Steemit | YouTube

artificial intelligence, chess, games, puzzles, Uncategorized

Computer-Generated Chess Problem 02419

A ‘KQBBP vs kbnppp’ chess problem generated autonomously by Chesthetica using the approach known as the DSNS from the sub-field of AI, computational creativity. Depending on the type and complexity of the problem desired, a single instance of Chesthetica running on a desktop computer can probably generate anywhere between one and ten problems per hour. The largest (Lomonosov) tablebase today is for 7 pieces which contains over 500 trillion positions. With each additional piece, the number of possible positions increases exponentially. It is therefore impossible that this problem with 11 pieces could have been taken from such a database.

8/3n2K1/5p2/3B4/5B2/2p3b1/1k2PQ1p/8 w – – 0 1
White to Play and Mate in 5
Chesthetica v10.82 : Selangor, Malaysia
2018.12.4 4:10:49 PM
Solvability Estimate = Difficult

Some of the earliest chess problems by humans are over 10 centuries old but original ones by computer are very recent. What was the machine ‘thinking’ when it came up with this? Leave a comment below, if you like. Collectively, these puzzles are intended to cater to players of all levels.

Main Line of the Solution (Skip to 0:35)

Facebook | Twitter | Book | Website | Reddit