Re:Ask HN: Why isn't email spam solved using Proof of Work yet?
by jfengel on Friday January 08, 2021 @03:27PM
(#25689978)
I think the answer is that machine learning solved the problem better. My GMail spam folder used to be in the tens of thousands from the last 30 days. Now it's in the low hundreds. (Low enough to locate the occasional false positive.)
I can't tell if Google is just not telling me about the vast ocean that the filter considers obvious, or if it's just gotten so effective that the spammers gave up. Whatever it is, it's working. I dunno if other email providers are similarly effective, but the tech exists if they want it.
Now, there's still web forms, which don't have the vast spraying power of SMTP, and also don't have the same kind of access to data to drive email spam filters. They usually try to de-automate the process with CAPTCHAs, which are also kind of a proof-of-work system (a "work" that's supposed to be cheap for humans and expensive for computers).
Maybe you could install a proof-of-work based system there? Not being email, it sidesteps some of the issues on the form reply that SI_Rob reposted.
I can't tell if Google is just not telling me about the vast ocean that the filter considers obvious, or if it's just gotten so effective that the spammers gave up. Whatever it is, it's working. I dunno if other email providers are similarly effective, but the tech exists if they want it.
Now, there's still web forms, which don't have the vast spraying power of SMTP, and also don't have the same kind of access to data to drive email spam filters. They usually try to de-automate the process with CAPTCHAs, which are also kind of a proof-of-work system (a "work" that's supposed to be cheap for humans and expensive for computers).
Maybe you could install a proof-of-work based system there? Not being email, it sidesteps some of the issues on the form reply that SI_Rob reposted.