| Types of problems to solve (and avoid) |


Some problems are better suited to get started on than others - in terms of how likely you are to succeed. This is especially important if you are short on some resources like time or finances or compute power.

As a rule, it is best to stay clear of what are known as "wicked-problems". These are, loosely speaking, ill-defined problems with a likely infinite solution space.

These problems can be quite enticing from the allure of solving-the-challenge or making-a-difference-in-the-society that they present. But that can be deceptive.

Even in case of well-defined, non-wicked problems, you have a large class of problems to chose from. But purely from the PoV of solvability, ultimately, you want to work on problems which are largely tractable. For such problems, the solutions can be either easily scalable or not so much without using a lot of computing power or automation.

Problem/solution space based on tractability and scalability Problem/solution space based on tractability and scalability

Identifying "wicked problems" and associated difficulties


A good paper to read up on the complexity of planning problems