Let's be blunt. Games are drugs.
His point is that most computer games deliberately generate an addiction response, and he's right. When I was in the biz myself people routinely talked about addictiveness as a positive good and a sign of quality.
I would argue that a significant dose of game addictiveness is part of the cocktail that makes a good game good. Last year I went through a Plants vs. Zombies phase, and the addiction response to my progress at learning the skill of the game was part of it. But the addiction of PvZ isn't the only good thing about it; there's also the charm of the art, the pleasure of developing an understanding of the game, the progressive reveal of little puns and jokes, and more.
Plants vs Zombies isn't quite great art, but it's very good art. I notice that my niece loves it despite being too young to actually understand the game. She's not having an addiction response, she's just enjoying the animations and the sensual satisfaction of the interactivity. And I submit that computer games are a medium capable of great art, though the examples are precious few.
So I don't want computer games reduced to and defined by their addictiveness. That's at least a waste of the artistic potential of the medium and at worst a real social problem. That's why I object to games of the Zynga model, which are designed to be not fun but nothing more than addiction triggers, as Ian Bogost tried to show with Cow Clicker, his game satirizing Farmville. (Jason Tanz has a terrific long Wired article about Cow Clicker, and Bogost mocks “gamification” with a recent post about cow-clickification.)
Unhappily, the addition factor is arguably the main force shaping the computer game industry. Cook offers a really illuminating discussion of the business and artistic implications.