We are all afraid, to a greater or lesser extent, of the enormous potential of artificial intelligence. Even tech visionary Elon Musk claimed that AI could turn us into his pets if we didn’t prepare. How not to fear him? This technological breakthrough has the potential to surpass humans in all practical aspects.
Solutions scientists propose range from controlled development to a fusion between humans and technology. However, a new approach has recently emerged that indicates that what is missing to reach the full potential of artificial intelligence – and to do so safely – is to develop more human qualities: empathy, self-awareness, and gradual learning.
This was concluded by Mark Lee, author of the book How to Grow a Robot. The scientist affirms that through two principles, it could be possible for artificial intelligence systems to communicate in a friendly way with humans.
These two principles are theory of mind and physical learning. The analysis concludes that a physical body is essential for the technology to create empathy. Friendly and lasting relationships between humans and artificial intelligence would be created through empathy. But why a physical body?
Related: Is Ai Smarter than Humans? Human IQ vs. Ai Benchmarks
Why would Artificial Intelligence need a body?
In psychology, there is a concept called “theory of mind.” It is the understanding that other people have their way of thinking and see the world differently. For this reason, when someone talks about her experiences, what they relate to is appreciated from her perspective, and meaning is obtained from her perspective.
When communicating with a computer system, the only intention is to converse. These are not conversations but script processing. For there to be a social interaction, a fluent conversation would be needed. In this conversation, artificial intelligence should remember the person, his aspirations, beliefs, and intentions from his words.
For an AI to achieve all of the above requires the parties involved to have a sense of self. This is where the need for a physical body becomes important. To understand someone else, self-knowledge is required, that is, a subjective perspective. This subjectivity must include points of view, an awareness of one’s own space and knowledge of one’s abilities.
This means that artificial intelligence needs a physical body sense of itself. Artificial means of a body can produce concrete experiences, and only with these experiences will it be able to generate empathy with humans by understanding the differences between people. In the research, Mark Lee states:
How could a software box have a subjective point of view of, and in, the physical world, the world humans inhabit? Our conversational systems must not only be integrated but embodied.
Subjective Learning
This must be accompanied by a design that supports learning from subjective perspectives. To achieve this, we must base ourselves on our learning processes. We present a very large and rapid development progress when we are children. At this stage, we learn to control our bodies to perceive objects and environments. We also learn to interact and learn about the consequences of our actions.
Of course, this perspective is only accepted by some of the community of the computer world with further ado since much has been argued about the complexity of simulating the human brain and mind. However, it is a proposal that has interest.
If artificial intelligence and robotics developed similarly, they could learn like children. This would mean the entire learning process would be organic, progressive, and subjective. In this way, they could gradually generate more complex interactions that help them create more empathetic social interactions with humans.
