The Feynman lectures are obviously brilliant but think the computation lectures are probably a better display of Feynman's brilliance. It's quite stunning how up to date they are.

Although that being said the rough outline of a field is usually worked out almost immediately after a consensus forms that it's "real" so to speak.

The theory of computation hasn't changed a whole lot since those times - and feynman explains it very well to a laymen audience (which is what makes it great, as it's not filled with jargon).