Academic Earth

Academic Earth is a website that gathers thousands of video lectures from the world's top scholars.

From their FAQ:
"What is Academic Earth?
Academic Earth is working to bring together the best available content across many subjects and distribute it in an engaging and useful way to learners worldwide.

How many videos are available on Academic Earth?
As of this writing, there are over fifteen hundred videos from MIT, Stanford, Berkeley, Harvard, Princeton, and Yale. We are adding more every day. Some of these are grouped into courses, which each contain anywhere from 4 to 50 lectures.

Can I get credit for these courses?
No - but feel free to learn as much as you'd like.

Can I download copies of lectures?
Yes, when the "Download to iPod" option appears you can download a QuickTime version of the lecture. Just right click on the "Download to iPod" link and select "Save Target As..." or the equivalent for your browser.

Can you skip ahead during lectures?
You can skip ahead in lectures using the progress bar at the bottom of the video. However, right now, you need to wait for the video to load before you can do this. (Watch the dark grey bar). We're working on improving this for future releases.

How do grades work?
Lectures start out with a grade of B. From there the grade is an average of the grades that our users have given it. Course grades are based on the grades given to the lectures in that course. Highly rated content will show up first in browse results and in the Top Rated sections in our homepage. The more people provide feedback, the better the ratings will get."

Sample lecture by Guy Kawasaki:

The blurb for it:

"Kawasaki thinks that companies like Nordstrom, Audi, and Nike make the world a better place. His advice: don't start a business because you think that's what's hot. You should study and do what you love, he says. He can't promise you that the money will come, he does promise that if you start a company simply for the money, you will probably end up miserable."

Enjoy!
(From Metafilter)