Computer-Generated Chess Problem 03040

Consider this 'KQBB vs krbnp' four-move chess construct composed autonomously by a computer using the DSNS computational creativity approach which doesn't use any kind of traditional AI. Chesthetica has the creative ability to compose positions that may otherwise take centuries to arise in an actual game, if ever. 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 9 pieces could have been taken from such a database.

image.png

1K6/6p1/b4Q2/8/r7/4k1B1/8/2n2B2 w - - 0 1
White to Play and Mate in 4
Chesthetica v11.89 (Selangor, Malaysia)
Generated on 4 Sep 2020 at 5:20:29 PM
Solvability Estimate = Moderate

Even with the same version number, each copy of Chesthetica 'evolves' and performs somewhat differently over time. White is significantly ahead in material. Do share and try out some of the others too. Solving chess puzzles like this can also help improve your game.

Solution

<| Amazon | BitChute | Minds | Reddit | WordPress | YouTube |>

H2
H3
H4
3 columns
2 columns
1 column
Join the conversation now