"The most important thing to consider," I began, "is that our own internal research shows our competitors are beginning to approach Google's level of quality. In a world where all search engines are equal, we'll need to rely on branding to differentiate us from everyone else."

The room grew quiet.

I looked around nervously. Had I said something wrong? Yes. Not just wrong, but heretical to engineers who believed anything could be improved through iterative application of intelligence. Larry made my apostasy clear.

"If we can't win on quality," he said quietly, "we shouldn't win at all." In his view, winning by marketing alone would be deceitful, because it would mean people had been tricked into using an inferior service against their own best interests. It would be nobler to take arms against our sea of troubles and by opposing, end them.