I see two aspects to the 3rd rule: one, pick a problem that isn't currently popular with most people in your field - it may be so in the future, thus you need t

I see two aspects to the 3rd rule: one, pick a problem that isn't currently popular with most people in your field - it may be so in the future, thus you need to think ahead, OR, you need to look at a "solved" problem that really isn't, something from the past which you have new insights on. Avoid just chasing the current fad(s), whether in science or business. Two, develop skills that are unusual, relative to others in the field. For instance, in older engineering fields (civil, mechanical, industrial), there are many problems that could be tackled differently and/or better by someone armed with both engineering domain knowledge, and solid coding skills. But most engineers are terrible coders. And most coders know little about non-electrical, non-computer engineering.