Humans use objects creatively- a trademark of their intelligence !
Humans use objects creatively. The statement in itself talks volumes about the creative and the analytical aspects of human intelligence. Now, compared to other life forms on earth, humans certainly have an edge over them. And this alone becomes the very reason that humans are and have been highly inquisitive about their surroundings. They have faltered, fell, fidgeted, and fumbled to only conquer summits. And this thus makes humans, an apt species that to a great extent, befits Darwin’s, survival of the fittest theory.
A research was conducted at MIT’s Center for Brains, Minds, and Machines only to find out the underlying reasons for such human capabilities. The research was published in the Proceedings of the National Academy of Sciences. Researchers like Kelsey Allen, Kevin Smith, and Joshua Tenenbaum were curious to know about how humans use objects in creative ways? Thus, to demystify this tool-using ability, they made simple daily life examples that are oblivious to us, as a base for their study. A simple instance of replacing a hammer with a flat surfaced heavy object to tap-beat a nail into the wall. Such examples clearly show how humans use objects in creative ways, which is a trademark of their intelligence.
The Virtual Tools Game:
The team designed a game called the virtual tools game to observe the tool-using capabilities of humans. It required people to pick a tool from a given tool-set to complete a task. The task like putting a ball into a container was simply a virtual two-dimensional computerized scene. Now to solve such games, it does require prior planning, gauging, and then making a calculated move to launch the ball. It is this planning and gauging, which the researchers are trying to tap into. So, to ease out the study, they hypothesized the number of capabilities that people use to execute a certain task(or game in this case) and divided them into three.
Anyone who tries to solve these virtual puzzles will utilize his/her capabilities in three ways. Firstly, there exists a certain belief that one’s actions shall have an effect on the current scene and create a difference. Secondly, there is an inbuilt mechanism to think what would be the possible outcome of a particular action. And thirdly, an immediate update of the belief about what is the plausible action to get the required solution.
Thus, the team built a model called “SSUP” that is Sample, Simulate, and Update, which is an interesting piece of research. And the SSUP was made to play the same game as humans and the results came out to be surprising. Researchers found out that the SSUP model cracked puzzles at comparable rates as humans. And in similar ways, as humans did. While on the contrary, another model that played Atari games efficiently, was not able to solve puzzles. It did not solve puzzles because it did not have direct expertise or training on the same.
Humans use objects creatively- reason to study?
Now, the study is not an exercise in floccinaucinihilipilification rather a strong base for further research. The study of humans’ cognitive abilities, can be of major help in the field of artificial intelligence as well in the use of physical tools. It paves a way for innovative learning and modeling of new tools for a variety of problems. And, hence the study of how humans use objects in creative ways becomes very promising for the future.
What do researchers say?
Kelsey Allen, reveals thrillingly about their collaboration with researchers from different institutions. And also motivates people at the same time to incorporate and observe simple gaming problems in their daily lives. This goes for all the people from the cognitive community as well as other fraternities alike. He stresses the connection between the effects of using physical objects and the decision-making ability.
Trial and Error Learning
Now, Tenenbaum here highlights the contribution of such an approach towards imparting intelligence to machines. He talks about how the model of reinforcement learning is based on the trial and error method. And be made better by reducing the number of trials from billions to only a counted few. Thus researchers believe that by empowering reinforcement learning and reducing the number of trials and errors, the current study will only help us advance closer to new artificial intelligence approaches.
“Humans use objects creatively” is not just a statement but a cognitive approach and makes the base for broader research. It highlights the importance of incorporating games and tools in our daily lives that utilize our cognitive abilities. Models like the SSUP, reinforcement learning, and trial and error methods hold great potential for the future artificial intelligence field.