There are some well-thought out lists:
https://80000hours.org/problem-profiles/
https://www.ycombinator.com/rfs
However, something about them doesn't click. I don't think most of these are inspiring. It's best to solve problems that you personally face.
A lot of people run the calculations, and think that the most impactful thing they could do is help out some poor country. But they usually don't understand the root cause of those countries being poor. So they throw money at wells and education, when the real problem might be crime rate and gangs/corruption.
It's probably best to solve problems locally. Even if you're in the richest country in the world, there's probably a "backyard" that needs your help. I know people from small American states who would be grateful to not grow up as truck drivers or forced to work below minimum wage on a farm.
There's plenty of us who are in developing countries who understand the challenges of a developing country better too. Much of the problems in Malaysia/Malaya for the last 50 or 600 years has been a shortage of tech. The region has always been rich. Tech led to colonial conquest and later on, income disparity as the rich used technology to buy better mining/agriculture equipment and muscle the poor out. There's still a robber baron culture here. Discrimination is rampant - it's normal to get rejected for race, looks, or gender (especially if you're pregnant). It doesn't matter how much money you put into educating young women if nobody wants to hire women with degrees. And since people tend to copy "more successful companies", this culture is permeated.
It's funny to read HN and see comments who say unicorns won't improve the economy. Just the fantasy of having unicorns has improved tech and culture a lot. The power structure has shifted away from employers. Employers are giving much better terms, especially to disadvantaged people. So people are now able to live better lives, and it frees a lot of intelligent people who would otherwise be forced to work two jobs to survive.