Initialy invest all but 10M. Mostly safe and long term stuff like land. Set myself up with a perfectly sci-fi but economical life so I can dedicate the next 10 years to learning everything possible about neuroengineering and anything 1 or 2 degrees removed.
Get into a career to reverse engineer the brain... Depending on how things play out I'll start a pet company like Flugwerk and find some way to inject a ton of warbirds in private collections for as cheap as possible.. For people who'll fly them, not just display them. Target would be as many warbirds in the air, not profit. I'd probably keep profits only large enough to keep the machinery running. I'd at least fab myself a Ta152.
I'd never dip investments under 50% of the total sum. I'd only lend money to people who had a solid plan for profit and agreed to 25-50% returns to me in the first part of their plan but nothing after that, depending on whether they do it just for their own profit (50% ROI to me) or if they have some altruistic plan that pushes human progress (25% or whatever low ROI is worth it, in that case), like an angel investor of sorts.
Fusion, self-replicating machinery and nanomanufacturing, artificial intelligence, medical progress, space exploration (mostly consequent on progress in materials, AI, etc, I'd push bottom up, not compete with NASA or anything), and probably a good deal of effort in education either localy at a mid/high level, or as incentives for wider lower-level education for small kids... That last one would probably be a huge time sink, though. I'd probably not do it unless I could visit every target site myself at least once a month to see that things are working like they're supposed to.
