Chess computers have become increasingly strong in recent years. They have become so strong that not even the top players in the world can defeat the top engines. This can cause chess players, especially club players, to think that the best way to learn chess is from chess engines. This is far from the truth.
Even if chess engines always suggested the best moves, they would not be the best way to study chess because they are unable to explain why. If you don't understand why a move or strategy is good, how will you be able to apply it in your own games?
The main reason why chess engines are so strong is because of their calculation. Chess engines can calculate millions of moves per second, far faster than the top players. While their positional understanding can be lacking, sorting through millions of positions can make it accurate enough to rarely blunder or miss tactics.Coorispondance chess has proven that players can understand chess better than engines, a human plus an engine against another engine almost always defeats (or draws, but rarely loses) to another engine on coorispondance chess (see servers where engines are allowed, such as ICCF, for more information). This is because the human's engine can detect the other engine's tactical ideas, leaving the player room to positionally outplay the engine.
I recomend players below 1800 USCF to analyze their games without an engine. If you are over 1800 USCF, you might find some value in training with an engine to practice theoretical endgames (winning won positions) or analyzing sharp opening variations with an engine. However, if you use an engine with your opening preperation, do not let the engine do all of the work for you. Engines generally understand less about the opening than theory. Chess engines can be useful to test your ideas (novelties) and understand concrete potential at the end of a complex and tactical opening variation.Don't be afraid to test your ideas against the engine's recomendations, be persistant! This can not only help you understand the opening better, but the engine might eventually agree with you.
"Serious creative work by the leading players on mastering the middlegame lasted for a comparitively short time - until the appearance of powerful computer programs, which largely hindered that positive tendancy." "The intervention of computers has damaged the study of typical positions, since a machine has a poor understanding of ideas, whereas it understands very well what material is." - GM Dragoljub Velimirovic
"In my opinion, the players are the main problem, not computers: the players have become too cautious! In former times, on encountering an opening suprise, they made every effort to refute it over the board. But today a self-preservation insting operates, a fear of the opponent's computer preperation. Players are afraid to make a mistake in their calculations, especially when a sharp reply is required." "Every player knows that the threat is stronger than the execution, and the threat of the computer is stronger than the computer's move! I have to be hopeful about the future. Eventually there will emerge a new generation, which in the end will overcome it's computer phobia and forget its own weapons." - GM Andy Soltis