Ask for forgiveness is to prevent you getting stuck into analysis-paralysis (as much as it grosses me out using business-speak like that, the term fits).
The top comment here as of now, about exposing a firm to $5b worth of risk?
C'mon... Do really think anybody is actually advocating that?
Asking permission might not be the right answer but checking your ideas out definitely beats just going for it.
However, by default I'd say there are great ideas, and if they got the boot for asking forgiveness, maybe it's a good thing for your friends - as it it's better to be working where one's value is recognized, than at a company digging its own grave in the midst of a lack of feature, lack of ambition and lack of insight.
The article mentions it only works if you hire insanely smart people, but that's also not a sufficient condition. You can have really amazing programmers, but if they don't have a strong understanding of your demographic and your goals, they're not "smart" in the way that's most relevant to this particular sort of decision.
I also dislike the phrase to begin with. Asking forgiveness rather than permission is often more effective, but that doesn't mean it isn't also irresponsible and a breach of trust. In an environment that doesn't claim this sort of non-regulation (i.e. most of them), if your integrity sells for the value of one commit and the risk of breaking or misdirecting your employer's project, you have more fundamental problems than potential bugs in production.
I'm not sure of the full intentions of the article nor your comment, but I think you may have missed a major point here. The article does not seem to argue that insanely smart people know all the right answers; instead it argues that insanely smart people have a pretty good idea when something is a bad idea, and will be the type of person who works out the right answer prior to pushing it live. That said, the other cases is covered as well...
Actually, mistakes are fine. They’re something you trade off for other variables like speed of iteration.
Yet another pretentious startup blogger.
I am insanely smart and capable, yet for anything beyond a cute web app, I don't think you should use users as a debugging tool
OTOH I don't accept he sentence but that is a different question.
Employees get to pick their own projects: their work model is based around self-gathered work groups. Got an awesome idea? Go make a group, gather fellow employees and DO IT! The company is so centered around this fact that employee's desks have wheels so they're easier to move around.
Okay, this is just anecdotal evidence and they might just be the luckiest company ever, but they managed to pull the trick and it would be dumb to overlook their success.
Here's a cool handbook for new Valve employees if you want to take a deeper look at how Valve works:
Dear Board of Directors,
During the past few weeks I have been engaging in unauthorized trading. I bet that the Euro would weaken drastically, hoping to make the company at least $4.6 Billion. Instead, we're 5 billion down, as the situation changed.
I was going to ask for permission to use all that cash to trade, but I read on a blog that I should ask for forgiveness, after the fact. So, forgive me.
Remember, a hacker is not Google or Apple to get away with book scanning ("ask us nicely to remove scans of your books") or location tracking.