Take a look at this 'KQRBN vs kqrbnn' five-move chess problem generated autonomously by Chesthetica using the DSNS computational creativity approach which does not use any kind of machine or deep learning. Chesthetica is able to use the technology to express original creative thought in this domain. It also never had behind it a team of highly skilled programmers, consultants or the kind of hardware that millions of dollars from a company like IBM or Google could buy. Chesthetica is able to generate a variety of compositions using any combination of pieces. Read more about it here. Any chess position over seven pieces would likely not have been derived from an endgame tablebase which today is limited to seven pieces. Work has only recently begun on one for eight pieces.
White to Play and Mate in 5
Chesthetica v12.64 (Selangor, Malaysia)
Generated on 14 Dec 2022 at 7:09:42 AM
Sometimes an earlier version of Chesthetica is credited for a more recent problem because that version was still running on that computer at the time. White actually has less material than Black yet still wins. The white army is down by about three (Shannon) pawn units in value. Do you think you could have composed something better with these pieces? Share in the comments and let us know how long it took you. Collectively, these puzzles are intended to cater to players of all levels. If you're bored of standard chess, though, why not try this?
Solution
<| Amazon | Minds | Reddit | WordPress | YouTube |>