Computers are excellent at solving problems… that is as long as the problem can be reduced to an algorithm (artificial intelligence researchers are working to change this, but most folks who use computers solve such reduced problems with them). Reduction to an algorithm requires rules to be clearly and completely defined. If the problem reflects the real world and then the cause and effect relationships must be known as well.
Algorithms are also context independent, so they can be used for every situation in which they are appropriate. Trying to figure out how to drop “stuff” out of airplanes and have it land on target? You may use the same algorithm whether at war or a humanitarian.
The most challenging problems, and the ones that should be the greatest focus of curriculum but that seem to the the least focus, are those that cannot be reduced to algorithms. This may result from them being too complex to accurately identify cause and effect. This may be because these are problems we did not recognize or that did not exist previously. Solving problems with computers is easy. Solving problems that we don’t even agree are problems, that is hard.