What is Cognitive Science?
The definition of Cognitive Science — scientific study of mind and it’s processes. Often also called Cognitive Sciences, as it includes Psychology, Neuroscience, Artificial Intelligence, Linguistics and other disciplines. Cognitive approach is focused on the study of structures and processes in those fields in precise terms. The main object of study in CogSci is nervous system. By the way, the history of CogSci is closely related to the history of AI and it appeared around at the same time in 1950s.
It studies processes of decision-making and rationality, learning, attention, communication, perception, structures of neurons, brain, memory and language. Basic research methods include behavioral experiments, brain imaging, recordings of electrical activity in the neurons and computational modelling.
Theoretical research in CS is trying to connect the dots between all those fields and their findings. Unfortunately, very few theoretical models received significant support, so at the moment there is no solid common ground for Cognitive Sciences.
Among the most influential works in applied CS are Artificial Neural Networks for AI, Decision Theory for Economics, Politics and Sociology, Cognitive Behavioral Therapy for Psychology and Psychiatry, Brain Stimulation and Imaging devices for Neurology, human interaction with computers for designers, language comprehension models for Linguistics and more common learning theories for Education. It’s hard to overestimate how much this knowledge transformed the modern world. And today CogSci is growing at an incredible rate.
Current Trends
Artificial General Intelligence — a machine with general reasoning capabilities. Right now Reinforcement Learning with Deep NN and Multi-Modal NN are the most promising directions to AGI. Pools among researchers in the field show that on average they expect to see first AGI in about 30 years.
Brain-Machine Interfaces — direct connections between neurons and computers. These devices need to be able to decode signals from neural spikes and encode messages back into electrical impulses. Simple prototypes are already helping people to control prosthetic hands and feet. In the meantime DARPA and Neuralink are planning to create much more powerful interfaces.
Brain Imaging — set of techniques that aim to capture the structure and neural activity of the nervous system. Those images help us understand how information is processed in the brain and how diseases disrupt it. Their accuracy have become so great that some artists are using them in their work:
While multi-billion international collaboration Human Brain Project attempts to create a detailed model of the whole nervous system, less ambitious project has successfully “uploaded the mind of the worm into Lego robot”.
Overall, we are getting closer and closer to understanding of our brains and minds. And with it our abilities to create “thinking” machines are getting much better.
Originally published at Cognitive Chaos.
