WARNING: some it-who-must-not-be-named will be discussed.
As much as I hate to deny it, the it-who-must-not-be-named is influential. I would be living with my eyes closed if I said otherwise. To point the obvious, the nature of many jobs will change including research:
- Many jobs now becomes a validation engineer, including parts of research
- Research farms of agents will be undoubtedly be used to explore more ideas than imaginable.
While a lot of these loops can be done quite quickly, the limiting factor is reality and physical reality. It can simulate thousands of times in seconds but those sims don't always match physical reality. Especially in areas where we are still formulating its behavior. For example, in plasma physics. Sure, it can create a hypothesis or a design and test them in simulation many many times but that won't mean it would necessarily work in real life.
I'm pretty sure by the end of my PhD, it-who-must-not-be-named will have yet become many folds smarter as it has during my undergrad. I'm sure PhDs are using it like crazy. So will the skills developed in a PhD even be useful in the world afterwards? Will I be able contribute to the world at large? Will I contribute to the body of knowledge at large? Will the knowledge or skills be useful?