Ask YC: Why not release your email statistics involving strangers to the public to reduce spam/uninteresting email to you?

[amichail]

The idea here is to release the topics of emails from strangers as well as whether or not you generally liked emails on those topics.

This should serve as a guide for others to determine whom to email about various things.

For example, if you want to talk to someone about topic X, you can perform a search to see people who generally liked emails on that topic from strangers.

[ebukys]

While it's a good idea in theory, it would ultimately fail if we tried to apply it to reality. Think about it. Ignoring the fact that we are releasing personal information (no matter how trivial), attached to our email address, how often do you personally recieve spam that is useful or interesting to you? If the answer to that question was anything other than "never", guess what? It wasn't spam. Spam emails are mass emails that generally want you to click on their link. This will most likely direct you to their site, ask you to complete other offers, populate your screen with popups, etc etc etc. Perhaps some people like completing 5 offers which cost them X amount of money, in order to recieve an Xbox six months later. Perhaps some people like being asked if their breasts need to be twice as big. Perhaps some people like their computer crashing repeatedly because of a virus they've infected their computer with by opening an email, going through a link, or downloading an attachment. Personally, I am none of these people. Spam is classified as such for a reason. They are emails that we don't want, and don't ask for, the internet version of junk mail--only without the tiny bit of irritation that comes with getting the latter, thinking of all the better things someone could have done with the millions spent sending out glossy paper that immediately goes in the trash. Either way, regardless of whether or not I personally like spam emails, someone has already implemented the idea of sending people advertising emails based on what they like: mailing lists.

[amichail → ebukys]

I'm not referring to that kind of email, but rather, I'm referring to serious emails that are sent only to a few people. You may not like receiving all such email, especially if you get lots of email.

Some people are more likely to respond to email from strangers on various topics than others. This is useful knowledge for everyone to know.

[jcl → amichail]

It's also useful knowledge for spammers. If they have a list of people who are interested in topic X, then they can craft an e-mail with a subject and body that is relevant enough to topic X that the people will be likely to click on a link in the message. The reason that spammers do not already do this is because it is too much trouble to track the topics associated with e-mail addresses and because the number of e-mail addresses easily found for a topic is too small to justify the expense of writing a custom e-mail. With your proposed service, it sounds like it would be easy to both track topic associations and retrieve large numbers of addresses.

Really, it sounds like you're trying to solve two different problems: (1) given an e-mail address, find relevant topics and (2) given a topic, find relevant e-mail addresses. A solution to either of these problems has the potential to enable spammers, if it is standardized to the point where address/topic extraction can be automated. In this regard, it's hard to beat the current system of listing your interests on your web page (which at least solves problem 1).

[xirium]

I saw a proposal for spam whitelisting which involved users publishing a hashed list of email addresses. This allows people to receive messages from friends of friends but without revealing your address book. To prevent the data being aggregated, each hash would include your email address in the form:

you@your.tld:friend@their.tld

This prevents spammers aggregating large lists and then rainbow cracking the hashes to find valid addresses. You can also bulk the list with bogus entries to reduce spam and to increase your ego.

[rcoder → xirium]

That's just a semi-public form of greylisting. Greylisting works fine, but a large number of spammers have definitely figured out how to get around it at this point.

[rcoder]

So, you're basically proposing that people voluntarily opt-in to a world-wide database of email address/topic-of-interest pairs?

No way. Absolutely, no f-ing way would I ever let my address get in there, and I would strongly advise against any of the users I support ever volunteering such information either.

If you're truly seeking interested parties to email about a topic, a few minutes spent on Google should let you find folks who blog, post to public mailing lists, or otherwise mark themselves online as interetsed in that area.

The only value to a fully machine-readable database of the sort you describe would be for lazy cold-calling salespeople and spammers/phishers...though the separation is really only one of degrees.