Ask HN: What behavioral biometric can be reliable for both auth. and anonymity?

1 points by scobar 9 years ago | 1 comment

Handwriting and gait are unique to each person, but seem difficult to naturally and consistently fake (hiding one's true biometric) if one did not want his/her identity revealed. They also seem difficult to use as authentication tools for our electronics in their current forms.

Internal timers as authentication seem like they could easily be integrated into our electronics, they don't have to be remembered like a password or pattern, and they can easily and consistently be faked (by counting to a different number for example). What other behavioral biometrics might share these properties?

I've written a short description of how I came up with the idea. You may find it interesting:

https://news.ycombinator.com/item?id=12551965

scobar 9 years ago |

While reading one of Richard Feynman's books, one of many curiosities he wrote about sent my mind racing. He and some acquaintances were trying to accurately count to 1 minute in their minds without help from any timing device. None of them got it right, but the interesting part is the result of performing the experiment multiple times. Each individual would finish counting at almost exactly the same erroneous time as they had each previous attempt, and the individuals' times were distinctly different than the others'.

At the time I was reading that book, I'd been wondering if there were better ways than a username/pass or physiological biometrics to authenticate someone yet would also allow anonymity when the user prefers. I thought this internal timer would be a great fit, but depending on its accuracy and if it's actually unique to each individual, it might be best when combined with other similar behavioral biometrics.

Note: I didn't even know the term behavioral biometric until a wonderful person whom I met at Startup School 2016 informed me. I appreciate that kind person, and I'm very glad I met him.