[room500]

That is a strawman argument. Of course WebView should update. The question is whether the structure of incentives contributed to a risky release.

If you "play the game", an error budget causes your releases to become riskier and riskier as the budget grows. This is optimal because a riskier release allows you to accelerate changes, allowing you to hit other OKR.

The issue is that this will eventually cause very risky releases (where you would slow down otherwise). In a sense, an error budget does eventually become "amount of intentionally deployed errors" because you are incentivized to increase risk until an error does appear.