Bets and Debts

16 July, 2024

At some simplistic level software (all?) management is about who gets to make bets and and who has to pay debts.

Making bets is great. You get a green field! Maybe it blows up and you get a ton of credit. These are new feautres. AI/crypto integration. New development paradigms. Everybody wants to make bets. And some bets turn out great. Hirign Rob Pike to make Golang. Letting K8s come into existinance, gmail, azure, aws were at all one point bets inside of big (or at least highly profitable companies). But most bets fail. It’s nice if they fail fast and conclusively but worst they may linger. Maybe it would have worked out if we’d thrown more people/funding/attention its way. But nobody actually knows.

In a startup it is all just a question of which bets to make and if none of them work out you go bankrupt. Oh well some VC loses his money and founder loses some rep.

Inside a mid to large company bets are tricker because you also have to pay debts.

All existing services are always accruing debts. Updating your dependencies. Keeping up wwith security. Improvng alerts. Smoothign out user and internal errors. Adding features that aren’t going to change the world but make using your existing product less like chewing rusty nails.

Nobody wants to pay debts but they just get larger over time and will either choke your service or if you deprecate you lose credit with customers. So they’ve got to be paid. But if you’re prett succesful paying debts as low level person your upside is pretty limited. If you mess up you actually bring down things people are using.

So everyone wants to make bets and not pay their debts. Best is when a bet actually creates debt. This usually happens by shipping somethign that is onl a middling success because “it just needs more exposure”. Oftent he bettor leaves before debts are paid.

Have a nice day.