What's to stop you from making use of that data?
I'm not attacking your character or suggesting you will make use of the data, but it's a pretty tempting pool of data to sneak a look into. "Oh, so-and-so at Twitter earns that much, interesting. Holy crap, I met this guy last week, I had no idea he got screwed so badly by his current startup. Huh, looks like startup X has a fair bit of money."
A lot of this is actionable data. If the startup has IPOed, you might even fall under insider trading regulations and need to make sure you don't trade on it. On a purely ethical level, you might want to consider what sort of level of access you should have to the data.
Finally, there is, of course, the possible issue that someone might offer you a tidy sum of money to buy that data off you, once there's enough of it.
I would not advise anyone to put their data in there as this stands.
The vision of Ackwire is to make startup compensation data transparent to prospective employees and employers.
Email addresses are encrypted and stored in a different database. It's true that they aren't hard to manually access, but this presents a reasonable barrier.
The same argument applies to any web service that stores personal information. I'm sure there are administrators at Facebook, Gmail, etc. who could read all your private messages, view your photos, etc. But we assume they don't, and most of us use these services anyway.
"A lot of this is actionable data. If the startup has IPOed, you might even fall under insider trading regulations and need to make sure you don't trade on it."
I'm not a securities lawyer, but this doesn't seem to pass the material non-public test. Significant shareholders in a public company already have to disclose their holdings, publicly.
"Finally, there is, of course, the possible issue that someone might offer you a tidy sum of money to buy that data off you, once there's enough of it."
I think this is jumping the gun a bit, but any buyer would have to keep the current user agreement intact (posts are confidential and anonymous) or explicitly ask users to opt into the new terms.
I wouldn't trust a random person with this data, but I'd trust a member of a somewhat-meritocratic community like hackernews or quora, university alumni association, etc. It's fairly clear why this data is being collected, what is being done with it, and my potential downside risk is fairly capped. At some point, being overly paranoid is too much of a cost.
Yes, a technical solution to this class of problem would be preferable to social trust. (and is one of the things I'm working on, actually...)
The questions you need to ask yourself is:
1) is there any potential downside whatsoever to submitting my data here? If this data is on the front page of the new york times tomorrow, and everyone I know knows it, do I stand to lose anything because of it?
2) is there any real upside to submitting my data here? (seeing other people's data has obvious upsides)
Unless 2 is clearly bigger than 1, submitting your salary info here is an irrational decision.
This could
been using it for years..
Second that.
If email addresses are stored 'encrypted' how can they be kept 'confidential', you store them for a reason or you don't need to store them at all, hashed or not.
What gain do you have from storing the hashed address?
The best credential would be something like "is a member of FB or LinkedIn with a sufficiently filled in profile" but nothing personally identifying.
For the traditional version of this (alumni association stats), you have a trusted third party (the association), or restricting access to people on 18.0.0.0/8 or who have kerberos credentials, although logging anything personally identifying.
Anyway, it's the easiest thing in the world to create a throwaway email account, so the data still can't be trusted.