[falk]

The biggest problem I have with this is that it attaches you to your friend's profile and whatever he does online. So, for example, what if a buddy of mine posts that he just smoked a joint, and then he endorses me for a job. Unfortunately most employers will likely draw certain conclusions from that, "He hangs out with those kinds of people? I'm not hiring him." It could be much subtler, like your friend posts "I love Bob Marley."

[jongold]

Absolutely agree — I think that's the most important thing we want to test with this iteration :)

I'd think that people would choose the professional connections they trust rather than their stoner roommate from college, but there's only so much hypothesising we can do without real people trying it out!

I'm not a huge fan of Bob Marley either, but I know what you mean. It could be your friend who is, say, a prominent Rubyist tweeting a joke about PHP as you're applying to Facebook. We'll see how that plays out, but I'd think that an 'old school' job reference would have the same risks.

[falk]

Awesome, I'm glad your really thinking everything over. To be honest, it wasn't entirely clear to be that you can pick the friends you want to tweet about you. I guess now looking at it that seems rather obvious, haha. It would be really cool to see a recommendation engine, for example, "We think you should ask your friend John Doe to endorse you because he has a massive Twitter following, speaks regularly to companies X, Y, and Z, whom are all hiring. We've analyzed his profile and it looks pretty non-controversial." It's good to see you guys are making use of Twitter and not Facebook. That's a huge selling point for me. Good luck to you guys.

[jongold]

The first few very, very early prototypes were built because I, then my girlfriend, then some of my closest friends were in the situation where they needed something like this - so I'm quite passionate about making it work :)

That's an amazing idea — we hadn't thought that far ahead yet in terms of NLP stuff, but we can definitely do some amalgamation of Twitter and, say, GitHub data to help you know who to target :)

[stefl]

It's a good point that you can't be responsible for the things that people who recommend you say online.

The idea is that you'd share the link privately just with people you really trust, so it's unlikely someone you don't trust would add their recommendation. I agree - you probably need to be able to remove any recommendations that you don't think are appropriate...

[scottmagdalein]

I love Bob Marley. I don't smoke pot.

The same correlation (or guilt by association) could be made by the people you follow on Twitter or your Facebook friends or the people who endorse you on LinkedIn.