I don't know enough about any of the relevant fields to speak with confidence about what advancements are more likely to happen in the next few decades, but it

I don't know enough about any of the relevant fields to speak with confidence about what advancements are more likely to happen in the next few decades, but it feels that the number of things we'd need to get better at to send humans further than lunar orbit with a decent chance of survival represent a far greater challenge than the relatively modest progress we would need to make in AI to make more capable and independent exploration craft.