Posting Anon because, well, I hope it’s obvious.

There are only a small subset of things you can burn out people in building. And all of them are (at time of building) trivially inconsequential.

Let me give you some examples that my team and closely related ones are working on.

An infrastructure for a national Financial Services sub-industry. If it falls over, so does that industry within a day or so.

The entire technology ecosystem by which a national chain takes in money. Kill that, kill the company.

An infrastructure which, if it fails, takes the global financial system down in under 10 minutes.

The systems that run all billing for a global telco.

The systems that enable much of sub-Sarahan Africa to make non-cash payments to each other. Paypal’s a tiny player by comparison.

When you’re working on things that really put a dent in the universe, you cannot, cannot have the people building them anywhere but right at the top of their game.

There’s a reason for the 40 hour week y’know – people aren’t productive working much more than that for a sustained period (read: more than about 3 in a row). After that, it’s just machismo, but the measurable output doesn’t really change much, despite the bullshit about ‘Internet Time’.

Chucking out a lightweight demo that works like crap (but better than nothing) is possible under that pressure. Doing it right isn’t.