Contemplate this ‘KBNNPPPP vs krp’ four-move chess problem generated by the program, 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. 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. Noteworthy here is that a chess position with over 7 pieces could not have been derived or taken from an endgame tablebase because 7 pieces is the present limit.
2N5/1K2B3/8/2N5/3k2Pr/5P2/1p1PP3/8 w – – 0 1
White to Play and Mate in 4
Chesthetica v10.82 : Selangor, Malaysia
2018.12.15 2:58:03 AM
Solvability Estimate = Difficult
Some of the earliest chess problems by humans are over 10 centuries old but original ones by computer are very recent. Chesthetica composes everything autonomously (no human intervention) and even chooses the main line of the solution to show you. Try to solve this as quickly as you can. If you like it, please share with your friends. Over time, the tactics you see in these puzzles will help you improve your game.
Main Line of the Solution (Skip to 0:35)