Oxytocin: trust behavior between human and machines

Suspicions of <Oxytocin increases trust in humans(Michael, 2005)>

https://www.nature.com/articles/nature03701

Today I read one paper:<Oxytocin increases trust in humans(Michael, 2005)>, and want to put forward some of my suspicions and extend the trust behavior between human beings and machine.

The paper designed one trust experiment and one risk experiment to compare the subjects’ responses towards the monetary rewards, the key control hormone is ‘oxytocin’ . Oxytocin is a hormone which is correlated significantly with our social perception, eye gazing behavior and recognition to facial expressions.

I think this trust game’s idea is more like game theory and prisoner’s dilemma. The author wanted to show the correlation between utility of monetary rewards and trustee’s trust behavior. As we can see in the figure below, every time the trustee pays back more money than previous transferring, apparently the trustee should behave more selflessly to proceed the idealistic model. The authors design the subjects interact with people in trust experiment, while interact with computers in the risk experiment. They supposed that people will have less pressure when interaction with computers, thus the oxytocin will show significant change.

Michael, 2005

The result of the experiment is no significance between control group and treatment group. But I think the discussion should be focused on the designing improvements.

Firstly, how can we assume that those people will show more trust to people rather than the computer when predicting the monetary rewards? If we trust people, but the people show their selfishness, then it is a prisoner’s dilemma, each part will show their selfishness, and then the transferring amounts will go down, or they just maintain the original price when transferring.

Secondly, oxytocin shows significant binding to brain limbic system: like amygdala. It is for sure that the consistent state of minds will reduce the anxiety in the social interaction. However, in this experiment, if the computer shows more consistent status, isn’t that the risk experiment will show the higher significance than the trust experiment?

Thirdly, when the authors choose the subjects, did they consider about their different social traits when doing the monetary rewards experiment? People show different incentives when they are in 20s, 30s and 40s.

The three suspicions may induce why the result of the experiment is no significance.

I think it is a good idea to explore oxytocin and trust behavior, and a good direction for adding computer, however, we should firstly explore subjects’ attitudes towards the people and machine. Actually, the the variation of attitudes can affect the results.