Friday, October 12, 2018

Outrunning the Herd




No surprise here - children are more susceptible than adults to influence by robots, even if they're obviously not people.

I'll paraphrase - a bunch of kids were put in a room with a bunch of robots and asked to do dumb sh** that they wouldn't or shouldn't be doing. They could be influenced to do said things as long as the robots made it seem ok.

We used to call this peer pressure. Now we refer to a well-known scientific experiemnt called the Asch paradigm, and it is a standard for measuring conformity. If enough people believe something to be true, it starts to become "true," whether it's true or not. This is how society works. So this idea of succumbing to peer pressure isn't new or unknown, but the effect on kids vs adults, and that it's instigated by a  humanoid, is new, or at least a reminder of how susceptible we are to the influence of others, especially at a young age.

Our allegiance to non-humans shouldn't be too crazy to consider. In case you don't remember your childhood, or don't have kids of your own, let me remind you that for an entire decade in America, a 6-ft tall, purple fuzzy dinosaur told kids what to do. And before that, I do recall letting 16-bit video games take control of my life  for a good 5 years. Kids are not too picky about who they take authority from. If they did, puppet shows might not exist. Lacking a mature brain to generate their own purpose in life, they put a premium on the opinions, beliefs and especially the commands of others. The Bicameral Mind work of Julian Jaynes might also have something to add to this.

Next, I'm not sure how much this is related in the study-authors' work, but I have to bring up Hofstede's Cultural Dimensions, because one of these dimensions (individualism vs collectivism) measures the degree of conformity to be found in a particular population. Some cultures value conformity more than others, for example by forcing everyone to have the same haircut, and so this peer pressure effect may be even more pronounced in other parts of the world. This kid-influencing robot study was conducted in the UK, where there is a wider variety of haircuts.

Let's also take a look back at how Neal Stephenson was playing with this idea more than 20 years ago in a novel about an artificially intelligent book that was designed to teach children, and obviously to influence their worldview and hence their behavior.

And while we're at it, let's remember what has already taken place in America in regards to advertising to children. (I don't know enough about this to link something relevant; but I'm pretty sure there are at least some rules on how, when, and what can be advertised to kids, and it has something to do with McDonald's and breakfast cereal.)

Finally, the inverse of this study says that adults are not swayed so easily as children by a group of robots, and that we should take this as  a cool warning about such dangers in the future. But we can't deny that today, adults are already manipulated by robots*.

I can't say it better than the tech guru Jaron Lanier - social media is a behavior modification platform. We may not obey robots with the servitude of a child, but most of us tend to be influenced by our friends, and even friends of our friends, and so on.

*When I say "robots" in this context of social media, I'm talking about algorithms embodied by avatars that pretend to be actual human people, or the many shades of intelligentities that lie in between. 

Image source: Someone help me cite this X-ray Kermit picture? circa 2010


Notes:
Children 'at risk of robot influence'
Aug 2018, BBC News

Asch Test of Conformity - a test to measure the extent to which we are all suckers for social pressure.

Hofstede's Cultural Dimensions - offers a way to measure and compare some salient characteristics of culture

The Diamond Age: Or, A Young Lady's Illustrated Primer. Neal Stephenson, 1995.


Post Script:
What happens when you combine the Asch Paradigm and the Turing Test?


Post Post Script:
Would you look at that, just as we talk about how kids are in danger of being manipulated by robots, we get the first hints of how that might not be such a bad idea for adults. When you have an a**hole robot staring you down while you do your job, you do a better job. Go figure. I think we knew this already, with the whole cardboard cut-out cop reducing store thefts.

It doesn't matter if a thing is a robot or a can opener. We can't tell the difference. Forget the Uncanny Valley, if we assign agency to washing machines and our own mobile phones, what makes you think a crappy-looking robot isn't good enough to trick us? We are social creatures, our brain a social tool, and as such, everything is a person, and every act a social interaction. The way you handle your toothbrush has a wisp of sociality to it. Now make your toothbrush talk and have eyes, and try not to let it influence you. We can't.

Presence of 'mean' robot found to improve human concentration
Aug 2018, phys.org

No comments:

Post a Comment