Why Dreaming Electric Sheep Need Maslow to succeed
By McDonald, T. | Date 13th of December 2017
When a human wakes up in the morning, the next thing to happen depends on the needs of that human. For instance, if hungry, putting on a dressing gown and heading to the fridge is the most likely thing to happen. If the bladder is full, a trip to the toilet will occur before raiding the fridge. Furthermore, most people can perform these tasks independently, but what if the subject is a robot? Writers have written about autonomous robots and androids with super intelligence for decades. With that in mind, how would an autonomous robot achieve true autonomy? Before understanding an autonomous robot's needs, developers must first consider humans needs and second understand the meaning of an autonomous robot then apply human needs to the autonomous robot.
Maslow’s hierarchy of human needsFirst, before understanding the needs of autonomous robots, developers must first consider human needs. Maslow’s hierarchy of needs is a good place to start. After giving up on Law at Brooklyn University, Abraham Maslow went to Wisconsin University to study psychology where he came up with the hierarchy of human needs. The theory goes that the lower level needs must be satisfied before a person can think about the needs on the higher levels. For example, if a person is in a burning building they will not be worried about eating dinner. Furthermore, a person would avoid fire in order to protect him or herself from burning their skin or worse. Likewise, a mathematician cut off from society on a remote island with no modern resources would be more interested in finding food than working on a mathematical problem. In addition, the mathematician may wear eye protection in a sand storm. While the lower needs are deficiency needs, the higher levels are about needs of being. Deficiency needs mean lacking in something while things like aesthetic and cognitive needs are needs of being. The hierarchy starts with physiological needs at the bottom with safety needs, love and belongingness, esteem needs, cognitive needs and aesthetic needs on subsequently higher levels. Right at the top is self-actualisation, which is a state of happiness, worry free and being all that a person can be. Furthermore, the priority of a need can change. For example, if a person is happily creating a masterpiece of art and suddenly feels thirsty, the physiological need will jump forward in the queue of needs. The desire to satisfy that physiological need will override the aesthetic need. As a result, there is an order in which needs must be satisfied before a person can feel happiness. The hierarchy of needs is the ordering of needs in priority in order to achieve happiness and survival with the emphasis on survival and protecting vital organs.
Autonomous robotsSecond, what is an autonomous robot? In short, an autonomous robot needs no human help. This means that it can go about its business independently of human assistance. Looking at it another way, the robot must be able to make decisions about which direction to go or whether to change a route due to an obstacle or even how to best complete a task such as making breakfast. In order to achieve this level of autonomy, a robot will need to gather in information about the surrounding area or about the abilities of the obstacle. As an illustration, if the obstacle is a door, the robot needs to figure out which way the door opens or even if the door will open at all and decide, which tool it needs to use. Subsequently, to perform a multitude of tasks, the autonomous robot has a need for sensory input devices such as cameras, microphones, temperature sensors and actuators. An actuator is a mover that makes something work in a desired way and needs a power source and a control signal much like a muscle does if thinking of the muscle as the mover. Specifically, an autonomous robot would be a robotic device that is set in motion and then left to run and perform behaviours without the need for human assistance, bar some unseen event. Such robots are sense-think-act robots. To put it differently, if the robot can make a human breakfast without any need for assistance from that human, the robot is autonomous. If the robot makes breakfast without damaging itself or anything else, it is considered smart!
An autonomous robot must prioritise its needs.Applying the hierarchy of needs to create a truly autonomous robot. While humans have physiological needs and gain power from eating, a robot gains power from some other source such as electricity. Subsequently, the physiological needs of a robot may just be a need to plug in. In like manner, a solar powered robot will need to rest in the sunlight to recharge. Power is a top priority to a robot because if it runs out of power it will cease to function just as a human will cease to function without food or water. Humans avoid this eventuality by calculating its own need priorities. For this reason, the device will need to calculate how much energy it will need to complete a task or terminate a task if energy levels run low so it can recharge, or it will run out of power before the task is complete. Additionally a sense of spatial awareness allows a robot to navigate around obstacles; thus, it will need sensory input and connections to actuators. Humans have eyes, ears, brains, motor neurons and muscle in order to manoeuvre around and people are naturally protective of these. Following on from this, a robot may have cameras for eyes, microphones for ears, wires for motor neurons, CPUs for brains and actuators for muscle. With this in mind, it is obvious that an autonomous robot must consider its own safety needs or run the risk of breakdown, which in turn stops it from working independently. In the event that a robot finds away to complete a task, the ability to pass on that information to other autonomous robots would give an advantage to others. This could lead to a need for expression such as art; it may need to find the most elegant and efficient means of passing on the information to other robots. That is to say, it may not only need to complete a task, but also need to teach other robots the new trick. The other robots in need of a quick solution to an unknown problem would refer to the robot that has the best tricks. This may give rise to artificial intelligence in a new and exciting way! Nevertheless, the physiological and safety needs would come first. Unlike humans who prioritise needs in order to achieve a sense of self-esteem and happiness, a robot will not lose self-esteem or be unhappy if other robots do not accept it as a chief source of information or reject its program for poached eggs. An autonomous robot, however, will still need to prioritise its own needs such as power first, safety next then make a decision about a task while still tending to the lower needs in the hierarchy.
Understanding how humans prioritise needs to survive gives developers an understanding of what autonomous robots have as needs, but unlike a human, a robot does not have needs of being, yet. An autonomous robot is able to function independently of human assistance; consequently, it must calculate the priority of its own survival needs. As a result, autonomous robots of the future will have these needs and priorities built into their programming in the think part of the sense-think-act model.
Gross, R. 'Psychology: The Science of Mind and Behaviour', chapter 9 Motivation, fifth edition, [Book], Hodder Arnold, 2005.
Jet Propulsion Laboratory (1996) 'A description of the rover Sojourner', Jet Propulsion Laboratory, California Institute of Technology and the National Aeronautics and Space Administration [Online]. Available at https://mars.jpl.nasa.gov/MPF/rover/descrip.html (Accessed on 25/11/17)
I hope you enjoyed reading this blog either way please leave a comment.