Without being able to comment on your specific situation, I would strongly discourage almost anyone who wants to have a highly impactful career from dropping out of college (assuming you don’t have an excellent outside option).

There is sometimes a tendency within EA and adjacent communities to critique the value of formal education, or to at least suggest that most of the value of a college education comes via its signaling power. I think this is mistaken, but I also suspect the signaling power of a college degree may increase—rather than decrease—as AI becomes more capable, and it may become harder to use things like, e.g., work tests to assess differences in applicants’ abilities (because the floor will be higher).

This isn’t to dismiss your concerns about the relevance of the skills you will cultivate in college to a world dominated by AI; as someone who has spent the last several years doing a PhD that I suspect will soon be able to be done by AI, I sympathize. Rather, a few quick thoughts:

1. Reading the new 80k career guide, which touches on this to some extent (and seeking 80k advising, as I suspect they are fielding these concerns a lot).
2. Identifying skills at the intersection of your interests, abilities, and things that seem harder for AI to replace. For instance, if you were considering medicine, it might make more sense to pursue surgery rather than radiology.
3. Taking classes where professors are explicitly thinking about and engaging with these concerns, and thoughtfully designing syllabi accordingly.

Not an answer to your original question, but beware taking answers to the Metaculus question as reflecting when AGI will arrive, if by "AGI" you mean AI that will rapidly transform the world, or be able to perform literally every task humans perform as well as almost all humans. If you look at the resolution criteria for the question, all it requires for the answer to be yes, is that there is a model able to pass 4 specific hard benchmarks.  Passing a benchmark is not the same as performing well at all aspects of an actual human office or lab job. Furthermore, none of these benchmarks actually require being able to store memories long-term and act coherently on a time scale of weeks, two of the main things current models lack. It is a highly substantial assumption that any AI which can pass the Turing test, do well on a test of subject matter knowledge, code like a top human over relatively small time scales, and put together a complicated model car can do every economically significant task, or succeed in carrying out plans long-term, or have enough commonsense and adapatibility in practice to fully replace a white-collar middle manager or a plumber. 

Not that this means you shouldn't be thinking about how to optimize your career for an age where AI can do a lot of tasks currently done by humans, or even that AGI isn't imminent. But people using that particular Metaculus question to say "see, completely human-level or above on everything transformative AI" is coming soon, when that doesn't really match the resolution criteria, is a pet hate of mine. 

"EA has always bee[n] rather demanding,"

I want to clarify that this is a common but generally incorrect reading of EA's views. EA leaders have repeatedly clarified that you don't need to dedicate your life to it, and can simply donate to causes that others have identified as highly effective, and otherwise live your life.

If you want to do more than that, great, good for you - but EA isn't utilitarianism, so please don't confuse the demandingness of the two.