> If everyone who is secure is using a password manager 100% of the time, anyway, then you might as well use something stronger than a plaintext shared "word" (

> If everyone who is secure is using a password manager 100% of the time, anyway, then you might as well use something stronger than a plaintext shared "word" (a "key"). A related point I missed, too: given the assumption that everyone who is secure is using a password manager 100% of the time, the best way to encourage everyone to be secure, including average users, becomes "make it easier for, if not force, more average users to enroll in password managers". For better and worse, despite "no human on the planet can remember enough passwords to be secure" there's still a lot of human over-confidence that "we can do it", so making the case for password managers has been an uphill battle in some cases among average users. Passkeys present the opportunity "fix that bug" in password manager roll out. I know very few people with delusions that they can do elliptic curve calculations and/or remember large primes and can do large prime multiplication in their head enough to do private key math the same way that people think that they can remember passwords. The password manager stays the same recommended toolkit, but with passkeys there opens the opportunity to get the UX right and get even the most password manager skeptical enrolled into a password manager. Whether or not we succeed or squander this opportunity is still an open question, but that's some of the HN excitement about passkeys is that this is an opportunity to make the average user more secure for the good of all users. (A rising tide lifts all boats, as the saying goes.)