Summary:
"The human brain is really complicated, so I don't think we'll be able to do AI any time soon, because we need 'exponential' increases in software - also 'complexity brake'" (whatever those mean)
[edit] Changed TL:DR to Summary, which I should have used (or similar) in the first place.
Here's a great article by Paul Allen about why the singularity won't happen anytime soon. Basically a lot of the things we do are just not amenable to awesome looking exponential graphs.