Tech: Gaining trust on the internet

“Deciding whether or not to trust a person is like deciding whether or not to climb a tree, because you mght get a wonderful view from the highest branch, or you might simply get covered in sap, and for this reason many people choose to spend their time alone and indoors, where it is harder to get a splinter.”
Lemony Snicket, The Penultimate Peril


Trust is about building relationships. You balance the amount you commit to that trust against the amount of risk involved. All this is much harder on the internet, where you have neither personal contact, nor physical infrastructure to provide comfort. Other methods must be used.

It’s a big issue here in cyberspace. Luis Suarez pondered the issues involved in collaborative processes - how do you trust someone on the net to be an expert?

The trust process online doesn't translate too easily from real-world mechanisms. I was looking for a car recently, and there’s no real time to build relationships. The basis for trust I used was personal contact; I felt I could trust the green car bloke more than the red car bloke.* But I can't use this judgment process online. Those relationships either take time, or rely on offline experience and knowledge. On Ebay, the feedback mechanism uses other people’s testimony, other relationships built. Yet to be effective, sellers need to ramp up from scratch. How do you trust someone with no record? You build a record slowly, and people accede or not; the risks are greater where the record’s not there.

How do you trust a blog? That the writer a) is telling the truth; b) is saying something worthwhile; c) is worth spending the time reading? Again, it’s a matter of building up a record; sometimes with the help of links from trusted sources.

So far, then, I see three broad methods at work, when seeking to trust an unknown:
1) developing the relationship over time yourself;
2) recommendation from a reliable source;
3) recommendation from a large number of sources.
I’ve used all three in my time. Sometimes, more than one at once. For example, with Wikipedia I built up a level of trust over time (method 1) then examined its structure much more closely, which provided evidence for methods 2 and 3.

We’ve already seen variety of trust mechanisms build up on the net, and I’m sure there’ll be new ideas to come. But will they all make use of those three methods?


*In the end, I bought a car from someone else, on the basis of a better deal – and the fact that I was dealing with someone who I felt comfortable trusting (partly because she was a woman).