IEEE Future Directions
Related Articles
Selves and Emerging super self
We are clearly moving towards mixed teams, humans and machines operating hand in hand, learning from each other. The relation may become as spontaneous as the one experienced today in human teams (I am using the term “team” in a broad sense, including an aggregation of workers in the assembly line, in the design of a building, in a surgery room or in a classroom) where people interact with one another with some frictions here and there, but with the capability to overcome them, or it might become “intimate” like the one of a prosthetic limb with its human partner.
Like in human teams there are several selves playing the interaction game. The novelty is that some of these selves are machines’ ones.
As in the case of human teams there are some social aspects that need to be addressed. In a panel, discussing the future of work, a person from Intel involved in hiring stated that one of the characteristics that will be required in the next decades will be the capability to work in mixed teams, staffed by humans and machines.
I can appreciate this requirement. I have often experienced the frustration of having to interact with stupid machines, I can imagine that interacting with intelligent ones might be even more frustrating!
In a human team you often argue, negotiate and compromise. How can this be done with a machine, particularly one that knows it is absolutely right (could a machine realise it might be wrong? That is an interesting question…).
If a machine has extensive visibility and a sound protocol (of reasoning and execution) it is likely that it is right. On the other hand, right is not necessarily “right”, given other parameters that we take into account through our social intelligence. It may have happened to you that a better strategy in a team is not to enforce the right but compromise on what would be acceptable (provided it does not lead to major issues) because keeping a good general mood in the thttps://www.apa.org/monitor/2018/01/cover-social-robots.aspxeam will pay off at a later stage.
Will the selves of machine acquire this social intelligence? So far there has been little research work on the social behaviour of machines, because social was felt to be outside of a machine scope but the situation is changing rapidly.
Cynthia Breazeal, director of the Personal Robot Group at MIT said “developing robots for the home is the final frontier because designing robots that can interact with people in homes, schools, hospitals and workplaces is still a serious scientific challenge”.
Another important aspect in social relation is the development of a shared trust (sometimes of shared suspicion). Lack of trust is not good, but misplacing trust is also not good.
Cars manufacturers are equipping their quasi-self-driving cars with technologies to keep driver’s awareness awake. The few accidents that have happened with self driving cars have been related to an over-confidence by the driver that decided to stop paying attention to the road since the car was …so good! A few days ago, November 30th2018, police stopped a Tesla car on highway 101 in California when they realised the driver was sleeping at the wheel (it turned out he was drunk), an example of the trust people are starting to place on robots.
In a short while we will have digital twins interacting with one another and with us, and in the longer term they will also have some sort of “self” (selves). This is quite an unchartered territory.
Would these selves need some sort of social behaviour? Would selves mimicking our selves be able to avoid social behaviour?
As the ecosystem of selves will become more and more mixed and populated we are likely to see a super self emerging, as it happens with swarms.
Of course, the situation will be quite different from today since today we are external observers of a swarm behaviour, in the future we will be part of a swarm and our feeling towards an emergent super self might be awkward.
However, this is not something completely new. Today we already have this sort of super self emerging out of human swarms, or crowds as we usually call human swarms. This is addressed today in terms of emergent crowd behaviour but it is not very different from the self we have been discussing in the Symbiotic Autonomous Systems Initiative.
We are in for new scenarios as well as new challenges as machines and humans will interact more and more seamlessly with the participation of Digital Twins, a new player we still have to learn potential and impact.
