On @hitchBOT and trusting the kindness of strangers #hitchbotinUSA

A hitch-hiking robot has been damaged beyond repair, by person or persons unknown, two weeks into a trip across the USA having successfully travelled around Canada and Germany.  I have some identification with little #hitchBot after my own trips across the USA.  For those who don’t know the story, I spent the month of March 2011 travelling coast to coast across the USA.  I turned up in San Francisco on March 1st with a plane ticket back from New York City on 31st March and very little planned in between.  I then used blogging and social media to move across the continent through members of my online social network.  So the main differences are: I am not a robot; I have built relationships with people online over many years; I did not hitch-hike.

That last one is the most important factor, I think.  If I had insisted on hitch-hiking with total strangers rather than relying on the kindness of my existing network (albeit many friends-of-friends or people who didn’t actually know me well) I’d have had a very different experience.  If I’d been found in the woods outside a small town in Louisiana, dismembered and decapitated, quite a few people would have said “I told you so”  (because yes, they did tell me so).  I used to hitch-hike in Worcestershire in the early-eighties but pretty much everyone who gave me a lift back then told me that I shouldn’t be doing it.  It’s only got more dangerous and it’s always been considered more hazardous for women than men, let alone small robots with a limited vocabulary.  So I’m tempted to say that running an experiment like this is a bit like sending a young child out hitch-hiking.  Most people would look out for them, but sooner or later, they’d meet someone nasty.

Much has been made of the fact that it survived around Canada and Germany but the journey ended in the USA (particularly the irony of it ending in the “city of brotherly love”, Philadelphia).  Well yes, the irony is always there whenever any harm being done to anyone in Philadelphia, ever.  But that doesn’t stop it happening.  I’ve seen Americans beating themselves up for the fact it happened in their country and making this mean something about the USA.  I wouldn’t leap to any conclusions about national cultures.  Also I do think this was always going to be a time-limited project, the robot was defenceless and entirely dependent on who it met.  When I did my trip, I’ve no idea how many times I made a decision to go with one group of people rather than another, and I could always have talked my way out of trouble or run off even if my actual fighting, self-defence skills aren’t tops.  I felt as safe in the US as i have done in any European city and much more so than in some parts of some cities here.  So no, I don’t think it’s about the USA particularly – bad things happen to good robots everywhere.

I also doubt whether machines can build personal social capital within a human network.  They’re just too different.  Yes we think it’s sweet and its interactions through social media seemed to mimic those we have with people we don’t know very well.  It acquired some fans, though not quite enough to keep it carried safely.  I relied on some people who I had known for many years and I had also built up capital through many small interactions and bits of help of service.

The aim of the #hitchBOT experiment was to “see whether robots can trust humans”.  I don’t think that’s been disproved.  Trust is reciprocal, though that goes without saying in human circumstances – I trusted in my network not to chop me up into little bits in the woodshed and they people who helped me trusted me enough to let me into their homes and meet their families.  Trust comes from perceptions of your actions and your motivations.  For there to be reciprocal trust between humans and robots, those pesky robots are going to have to actively do trustworthy and helpful things for people with no expectation of rewards.