When a distinguished but elderly scientist states that something is possible, he is almost certainly right. When he states that something is impossible, he is very probably wrong.
Arthur C Clarke
Aviation is a strange mixture. TCAS/ACAS, ADS-B, SSR are all based on the 1940's IFF. In 1985 Boeing developed FANS 1 and CPDLC and uploadable flight plans into FMCs. Although standard for transoceanic flight, CPDLC
might be being used in en-route flight globally by 2040. Can anyone think of a communications system in normal use that is effectively the same as it was 40 years ago in the 1980's? Thanks to some mathematicians in the 1980's who felt formal proofs were required for DAL-A hardware such as FMCs, the current FMCs are based on architectures no more capable than the i286: Smart phones have far more processing power than the latest FMCs. There are grandfathered in concepts in the FMCs that can be traced back to the air traffic control concepts of the 1930's and similarly, the air traffic control concepts in use today have grandfathered in concepts from the 1930's and are not significantly different to those used in the 1960s. Some will say all this is due to the safety requirements - it is more extreme conservatism and this is 'good enough' why should we change this is the way we always did it. This cannot continue.
Autonomous aircraft of varying capability are
already flying, some are using Machine Learning type AI, some are now capable of air-air refueling and aircraft carrier takeoff and landing. Hypersonic flight is already possible and just like populations in the 1950's would not have considered transcontinental and transoceanic flight in less than 15 hours or flight from Europe to Australia was for the rich with multiple hotel stays, the idea of flight from USA to Europe in 2 hours or USA to Australia in 4 hours being the norm is something that will 'never happen'.
So we have conservative aviation meet rapid entrepreneurial 'get to market fast' development with Silicon Valley and others pushing autonomous and near space flight. I think that in ~15 - 30 years' time there will be several disruptive innovative changes. However many of the current 'innovative' ideas will fall by the wayside.
Your question - I would recommend that young children get an initial education in the STEM subjects,
"know your enemy" study AI, computing and telecommunications. Understand what computing Cloud services are and play with them (you may not realize it but you and I are using Cloud services now). Do the study at a university that has both STEM, engineering and flying courses there are several scattered around the US from Florida to the Dakotas. Students can graduate with ratings on twins and a 1st degree/masters in a STEM subject. This is the best of both worlds as someone said earlier students with a spread of expertise become far more employable.
Finally, children and students should look at learning in the way they look at learning a new video game - it should be fun and competitive.