Computer-Generated Chess Problem 02055
Take a look at this 'KQRBBNN vs kqqbnnp' study-like chess problem generated autonomously by a computer 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. 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 (such as using only a queen vs. rook and knight). Read more about it on ChessBase. 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 14 pieces could have been taken from such a database. Any analysis shown for this study could be flawed as chess engines may change their recommendations given more time. The first or key move, at least, is probably right.
Chesthetica v10.63 : Selangor, Malaysia
White to Play and Draw : 2018.4.28 12:04:09 PM
Some of the earliest chess problems by humans are over 10 centuries old but original ones by computer are very recent. White actually has less material than Black yet still manages to draw. The white army is down by about 2 (Shannon) pawn units in value. If this one is too easy or too difficult for you, try out some of the others. Feel free to copy the position into a chess engine and discover even more variations of the solution.
Solution (Skip to 0:35)