Honestly, as much fun as the Metal Gear series has been, I have to say one of its great failings post-Metal Gear Solid 1 (MGS2 and onward is where I say it has issues) has been not improving already-established concepts. So yes, Hideo Kojima does love exploring that which hasn't been done, be he doesn't ever change up that which is already in place.

In order to avoid making this a gigantic, boring post, let me just give two examples:
Metal Gear Solid A.I.: People can make the argument that they attempted to re-tune the A.I. or what have you, but the honest truth is I've never felt that they've changed much about the A.I.'s basic reactions to the player. It's always the same, and I've been able to just run through the motions in terms of avoiding (or shooting) the enemy verbatim since Metal Gear Solid 1. That's pretty sad given that there have been 3 games produced since.

Environment Interaction: For all the new attempts at "environment interaction" (see the OctoCamo in MGS 4 and Camo system in MGS 3), I still can't help but feel that this is still a real cop-out. With a stealth game, it's fine to have semi-linear gameplay, but when you restrict what you can/cannot shoot/mess with so that the player must use one of three or four well-designed and planned-out routes to attack the enemy or sneak around them, I fail to see where this is 'innovative' in an age where open-world gameplay is totally possible.

I'm not sure I understand the purpose of showing us the keynote speech (which, I note, had fairly poor filming and sound quality) just to have it cut off part-way through and not show us the whole thing.

Watching this 3 minute video hasn't exactly made me want to see the whole thing. Does the rest of the talk actually justify the incredibly pretentious title? By that I mean, is it anything other than Hideo Kojima ignoring what "impossible" means so he can talk about metal gear solid?