What kinds of trust can we build in online communities? If it is true that trust is a product of vulnerability, and that it takes some kind of risk exposure to produce trust, then how rich can our trust be here? And if one key ingredient to trust is a commitment to remain in a relationship (I can sacrifice your trust and suffer nothing if I'm willing to lose your relationship), then how do we overcome the obvious fact that online, there's neither judge nor jury... (nor police nor prison)...
posted by:
|
SF Bay Area
|
-
Re: Trust
Mon, April 5, 2004 - 5:43 PMThere are degrees of trust. There is trust based on a social contract and trust by probability. The most likely trust in social networks is the trust of probability. In a chat environment you follow along and notice a certain chatter offer sarcastic remarks to most of the posted statements. You can trust that when you make a post in the stream of conversation there is a high likelihood that chatter will respond in the same fashion. There is also the possiblity of trust built up over time. For instance in a tribe, By observing thoughtfulness of postings by an individual credibility (a form of trust) is developed. Future opinions and possible recommendations would be considered trust worthy because of the pattern. This would be an example of your latter form of trust. However you are right, unless our online actions seep into real life there is no police but is there also any need for them? As a norm we hold our online relationships at arms length hence preventing most types of personal/emotional injury. -
-
Re: Trust
Tue, April 6, 2004 - 4:15 PMKatie:
in your degrees of trust i think you actually have several different concepts. i'd break them down like this..
trust / confidence
trust > individual
confidence > operation, action, event, situation
opposite of trust is mistrust
the opposite of confidence however is uncertainty
for me trust applies only to people. i have *confidence* in my technologies, assuming that they're reliable, that's based on past experience and some amount of branding....
i agree that online relationships seem to involve less trust, but it might be because the risk we're exposed to is less. even if we're embarassed in front of the entire tribe online, nobody's going to see us blush....
the mix, of confidence in technology, and trust in people, would be an interesting combination and worth studying insofar as we're dealing here w/ both..
-
-
Re: Trust
Mon, April 5, 2004 - 8:25 PMHmmm...I think that this is a tangent, but how would people react to a SSS system where the trust relationships (and friendships, etc.) are defined by the participating individuals but those relationships are not themselves public?
I think that this occurs to me in part becuase it's a way to eliminate some of what makes me uncomfortable about social software, but it's an interesting idea anyway:
Upon joining the community you designate people that you consider friends, people that you trust, people that you think are interesting, and so on -- all in varying degrees; those statements are private to you, but they define how you see other members of the community. If you assign a high "trust" score to Ben, for example, and Ben has assigned a high "trust" score to Carol, then Carol will appear trustworthy to you.
Interesting thought, really -- the network is exactly the same for everyone, but perceived differently by each member. Kind of like life, I suppose... :) -
-
Re: Trust
Tue, April 6, 2004 - 1:49 PM>> a SSS system where the trust relationships (and friendships, etc.) are defined by the participating individuals but those relationships are not themselves public <<
I do believe that the way to go is via implicite relations not explicite declarations, just like irl declarations can be only a small part of the whole picture. -
-
Re: Trust
Tue, April 6, 2004 - 4:18 PMit seems natural that trust is an implicit aspect of communication, in fact one that's rarely declared but usually assumed...
-
-
Re: Trust
Tue, April 6, 2004 - 4:17 PMthis is tough. it's an interesting idea. i'd like to know if transference works here.
in my experience i trust only those whom i know. you could trust Katie for all the world, but if i entrust something to Katie based on your trust in her, am i not acting actually on my trust in you, not in her?
What application would come of this rating system? -
-
Re: Trust
Tue, April 6, 2004 - 5:05 PM> What application would come of this rating system?
Eh, I'm better on the contemplation than the implementation at the moment... :)
Right now I'm mostly just intrigued by the idea; I love, for example, that on a site implemented along these lines you would have no idea how trusted, intelligent, or friendly you were -- your "rankings" would depend entirely upon who was looking at you. You'd get a very interesting effect viewing others, as well: each direct relationship that you set up (A gets trust rating of 5, B gets a frienship rating of 4) would color how you percieve *everyone else* in the network as well...neat stuff, though I haven't really thought it out...
> In my experience i trust only those whom i know.
> you could trust Katie for all the world, but if i
> entrust something to Katie based on your trust in
> her, am i not acting actually on my trust in you,
> not in her?
Absolutely. Perhaps it's better to view this transaction as the foundation upon which your trust in Katie will be built -- perhaps you don't trust her yet, but you accept her as a good "trust risk" based upon my trust (not transference of trust, exactly). If the transaction were to be sucessful, you would (I assume) have some measure of direct trust in Katie, and some inclination to trust her independent of me. If the transaction were a spectacular failure, on the other hand, you would probably not trust me to the same degree -- and assign less value to my recommendation that others are trustworthy (again, not exactly transference, but something kind of like it).
Will think more about it. -
-
Re: Trust
Fri, April 9, 2004 - 12:33 AMHence we are back to probability,.. transactional trust. Can this be operationialized? I am thinking this might be a nice game theoretic not just a simple MLE.
-
Re: Trust
Fri, April 9, 2004 - 2:54 PMit's a very compelling idea and would make for a fun way to organize a social system.
one thing about trust. let's assume that trust is composed of parts information, parts future (time), and parts risk. My trust in X is a product of what i know about X, and any contextual information, how long i anticipate having future experiences/relationship with X, and what X has on me or could do to harm me.
my trust profile with X will be both general and specific. I may trust X to be honest and sincere w/ me, while at the same time i mistrust X's restaurant recommendations. Perhaps this is case of semantics: i trust the person even though i disagree with him/her.
there's another qualification that comes into play, and for us especially. Trust is qualified by what it is i am asked to risk. i dont think any of us is risking much here on tribe; the kind of trust we're talking about might even be best described as a kind of impersonal social investment and confidence -- expectations that members will neither abuse, expose, nor manipulate us. In this sense the members of an audience at a football game have as much trust as we do -- trust in the framework, the activity...
must think about it more... seems though that trust is both more and less than what we find in relations, esp the foaf relations under discussion here. it's transactional, it's generalizable, it's specific, it can be displaced or invested in social activity w/o implying particular persons, etc.
relations alone, and networks especially, resolve only one dimension of the trust problematic...
-
-
-
-
Re: Trust
Wed, May 26, 2004 - 5:05 AMJust a link for you: Trust Metrics Wiki bowie.itc.it/trustmetric...gi/FrontPage
-
Re: Trust
Thu, May 27, 2004 - 3:07 PMI am currently writing a press release for a company - when it gets released next week, I will share it with you...
I think is a whole new way to look at corporate trust, at least from a customer's point of view...I can't wait to share it with you - I think this company's approach to building trust is truly revolutionary.... -
-
Re: Trust
Thu, May 27, 2004 - 5:44 PMLooking forward to seeing it! post it to the group when you have it.
thanks
adrian
-
-
Re: Trust
Fri, May 28, 2004 - 2:23 AM"karma" was recently brought up in tribe ideas:
www.tribe.net/tribe/servl...iewThread.vm
which prompted ashton to bump the following thread:
www.tribe.net/tribe/servl...iewThread.vm -
-
Re: Trust
Thu, June 3, 2004 - 12:13 PMi havent had time to make it through the second thread, but i think i get the gist..
the issue i think is a simple one -- in any social system, an activity like this whose rules are explicit is exposed to gaming/exploitation. it would seem then that the activity collapses in on itself from abuse/misuse.
in the area of trust/karma/gifting, attributes are acquired as a consequence or byproduct of generous behavior. at least that's the way it seems to me. a gifting convention in which people recognize that the only reason they're giving gifts is to increase their status or to get them back in return, is either cynical or ceremonial. you find both. (it's commonly thought that a lot of our more "hollow" conventions are just the echoes of past economies, of times when tribute did in fact raise ones social standing, earn rights to a bride, appease ancestors, etc etc.
so i guess the issue is: can you take soemthing like karma and have it still be karmic if it has become completely instrumental/functional? I dont think so.
But something does happen -- and that's the part that we might be able to work with. In this case, what happens seems more like competition than it does generosity. So if social rivalry is a benefit to the community, the karmic function could be useful...
-
-
Re: Trust
Thu, June 3, 2004 - 12:48 PMdo you think a reputation system will still get corrupted if it's in a large network and is anonymous? (that is the "giver" is anonymous) -
-
Re: Trust
Thu, June 3, 2004 - 12:49 PMlike one of my sites, www.ratemyteachers.com - anonymous feedback... -
-
Re: Trust
Thu, June 3, 2004 - 12:51 PMthat's more of a one sided feedback system. you're less likely to have "give and take" in this kind of system. unless teachers promise good grades for good reviews... -
-
Re: Trust
Thu, June 3, 2004 - 12:53 PMActually, we receveive notes all the time about how this system has insitgated conversations between teachers and students regarding teaching styles, learning content, etc...
We've had positive impacts in thousands of classrooms - and thousands of thank yous too.. -
-
Re: Trust
Thu, June 3, 2004 - 1:01 PMi dont doubt the effectivity of the system. i'm just pointing out that it's different from a rating/trust/karma system on a social network in that it cannot have a "i'll rate you if you rate me" thing going on which would corrupt the system.
-
-
-
-
-
Re: Trust
Mon, June 7, 2004 - 1:24 AMI'm a technocrat but I have to agree, it is not possible to devise a mathematical model that will reward everyone with points exactly proportionally to the positive effect of their actions, there are so many problems with that that I wan't even list them here. All we can do is to devise a system that people will accept as fair enough - we have to pose a limit to the math and start psychology from that point.
A related book: "Punished by Rewards" by Alfie Kohn: www.gurteen.com/gurteen/gu...35005BC369/
I think it is all about 'internal' versus 'external' incentives. -
-
Re: Trust
Mon, June 7, 2004 - 5:46 PMI think for me the "rubber meets the road" where we have a framework that allows for trust relations between parties involved in commercial activities, such as designing and developing products online without "meat interaction". When the bandwidth arrives, video conferencing will allow some "talking-head" interaction, but a framework needs to be created that allows for interactions that will support commercial transactions. I like Adrian's analysis regarding the various aspects of trust and I think there will always be a "game-theoretic" aspect since no one can guarantee the future. Just because we have trusted each other for ten years doesn't mean someone won't betray someone else if the outcome looks good enough. Any system can be corrupted. But a metric, a rubric in which we can discuss tradeoffs, risks, uncertainties, plays, transactions, transaction metrics, possibilities to breach relationships, etc. A framework that incorporates and formalizes a sociology of interactional ontology. (I think I just had a "brain fart"). I am preparing to just start experimenting with an online work site and just see what works and what doesn't, and let trial and error be my teacher.
Gilton
-
-
-
