> There is no model where a 3rd party is involved that isn't pre-compromised, either legally (the government and civil actors via the courts), financially (stop

> There is no model where a 3rd party is involved that isn't pre-compromised, either legally (the government and civil actors via the courts), financially (stop paying, lose access), operationally (so sorry, your key expired), logistically (oh, sorry, our server is down. Can't load your keyring), or technically (um... this device thingy stopped working). Every one of these problems is really a showstopper when it comes down to it. Why? Because without perfect performance of third parties and/or hardware, the risk is your data being locked away forever (or even for a few days). This is what I'm saying about password managers. We *are already* in a situation where the best advice we can give each other is to get a 3rd party involved to store our passwords for us because we have to use passwords that we physically cannot remember as humans. Those same 3rd parties have the same complicated issues and responsibilities/trust legally, financially, operationally, logistically, and technically! Even how many of use Keepass or one of its forks: we accept some technical trade-offs ("It's open source!") for some massive legal ones. I tried to describe Keepass to my father and as an attorney he shivered in horror at the thought that there was no legal entity to hold accountable (through the courts if need be, through the Better Business Bureau and the court of public opinion at the very least) should things go wrong. My father wasn't wrong about that, and I've thought about that additional perspective for years now without good answers. That's also why I describe the above "choice" as an illusion of choice: our best security advice is that the safest passwords are already out of physical human memory (between minimum complexity requirements to avoid spray attacks and minimum uniqueness complexity to avoid reuse attacks) so remembering all of our passwords is just out. There are plenty of HN posts about algorithmic "passwords" that people can sort of memorize, but no algorithm can match the practical problem of the wide variety of different password policies across the web. The only real, practical option we have as a best practice is Password Managers. Which I want to make clear are already "3rd Party passwordless": and there we do have real choices right now between 3rd Party providers. I want to make it clear that just because I think passwords are broken I don't think choice is bad: I want us to understand as a group that the real fight isn't between "password managers" and "passwordless", because "password managers" are a workaround/hack towards "passwordless". The existing diversity in password managers is good, I don't know how we keep that in more "passwordless" systems, but I don't think we get there by championing for the "password" status quo. We can try to throw out the "password" bath water without throwing out the "password manager" baby, but mostly only if we start by agreeing that "passwords" are broken.