Research lessons from the making (and near-breaking) of Halo 2
"I'd argue that, just like for the players in our games, there is an ideal level of failure for researchers. If every study is equally successful, it just means that we aren’t innovating enough."
