Cognitive Science & Artificial Intelligence

Cognitive science is the scientific study of the mind and its processes. It draws from an array of disciplines, borrowing from psychology, philosophyneurosciencelinguistics, and anthropology.

Cognitive scientists study intelligence and behavior, focusing on how the nervous system represents, processes, and transforms information. Typical cognitive science analysis spans several organizational levels, learning and decision making, logic, and planning.

According to  The Stanford Encyclopedia of Philosophy, cognitive science fundamentally believes that thinking can best be understood in terms of structures in the mind” and the computational procedures operating these structures. The study of cognitive science is around 90 years old, beginning with psychology and the concepts of stimulus and reaction. Cognitive science was eventually backed by computer simulation models of human-level thinking. 

Artificial intelligence is a simulation of human cognition and intelligence processed by computers. AI assumes that machines can be programmed and trained to act as flexible and rational agents, perceiving the environment, and taking actions that will maximize their chance of reaching a goal successfully. AI enables the acquisition of information and follows defined rules so that a decision can be made, such as self-correction.

AI colloquially is applied to human cognitive functions, like learning and problem solving by machines.  As the abilities of machines grow, capabilities once thought to require intelligence are removed from the artificial intelligence definition. Optical Character Recognition (OCR), once considered the domain of artificial intelligence, has become a routine technology. Technologies such as playing games like chess and self-driving autos are still considered part of AI’s domain.

Cognitive Science and AI

The two fields of AI and cognitive science combine to understand the principles of intelligence, and to comprehend how exactly the mind learns. Initially, there was a branch of AI that emphasized the cognitive behavior of machines. However, with technological advancements, AI was allowed to encapsulate concepts of cognitive science, focusing on the ways humans, animals, and machines store information.

This switch led to the development of “intelligent machines.” These machines have made tasks such as speech and emotional recognition, planning, problem-solving, learning, and reasoning possible. The field progressed, and by using optimistic predictions, the invention of cognitive robots with a wide spectrum of cognitive powers was made possible. Such robots can perform open-ended tasks without human aid. They can learn and respond to complex situations. 

Sub-symbolic modeling is a cognitive science field consisting of neural network models and assumes that the brain is an amalgamation of simple nodes like a neural network. With this reasoning, the problem-solving capacity is believed to be derived from connections between nodes. This schema results in numerous approaches to depicting the mind with a computer, from creating artificial neurons to expressing it as a collection of symbols, rules, and plans. 

Cognitive Science’s Importance to AI

With AI immersed in several aspects of our lives, cognitive science gains importance. To benefit, industries need to understand the innerworkings of the mind and how AI will be programmed to please it. Nearly any field that has human-AI interaction will need to understand human mental and emotional processes. 

  • Educational organizations can adopt AI to teach more effectively, understanding the needs of each student’s learning type better. 
  • Tools for medicine and engineering should be better equipped to gauge and cope with the human user, simplifying things for them. Self-driving cars should still be designed for humans.
  • HR can use cognitive science to improve productivity and grow potential in individuals. 
  • Financial professionals providing automated services need to have a great understanding of the mind and how to appeal to it. The AI tools offered need to be simple to understand and use.
  • Apps need to ask the right questions so to understand customer preferences, combining identification like face or voice detection, and then know how to personalize the results and features to enhance the user experience.

AI and Historical Research

Advancements in technology make it possible for scientists to accurately simulate the human brain with computerized systems. In other words, the field of cognitive science has ensured the effective utilization of the power of computers to supplement the ability of humans to think. Computer simulation in AI can therefore be thought of as a reproduction of a system’s behavior, able to complete both simple and complex tasks and goals. 

It was believed that an AI’s ability to model the mind was questionable due to the lack of a computer’s ability to model language recognition. Several research projects of the past have been contradicted by today’s cognitive science. This result was due to an era where AI was used to understand intelligence in general and not that of humans.

Now, to gain further insight into the nature of human cognition, the goal is to develop machines with human-level intelligence. This has been the goal of the Turing Test and its variants. Failures have been seen with intelligent agents that face multiple situations with incomplete information. The basic encoding of data has become a limited approach to simulating human intelligence. 

The technological achievements in AI have widened the scope of the science to foster natural interaction. Handwriting recognition and speech recognition have attracted researchers, and this has expanded within the industry; 32 percent of executives consider voice recognition to be the most widely used AI in their businesses.

Cognitive Systems

SOAR Architecture

SOAR is a system of cognitive programming developed at the University of Michigan to simulate the brain. SOAR is termed an “alternative approach” because it stores and retrieves information from working memory. A process of reinforcement learning tunes the values of rules and creates numeric preferences supported by this cognitive system. To enhance flexibility, a structure within the working memory that considers or seeks rewards is created. 

The result is a simulation of virtual humans that can support face-to-face collaboration and dialogs.  While SOAR has integrated Natural Language Processing (NLP), emotion, action, and body control, it has been criticized for its lack of suitability for the real world, as opposed to the virtual world.

This is less of a concern as we move toward the Metaverse. With SOAR, there remains a question as to whether aspects of psychology must be minimized so that a better approximation of the knowledge level of symbol processing can be reached because SOAR attempts to replicate the evolutionary design process to create a better system. 

ACT-R (Adaptive Control of Thought-Rational)

ACR-R is a cognitive architecture that wishes to define basic and irreducible cognitive and perceptual operations. It’s inspired by psychological theories where every task a human performs is a set of discrete operations. It’s considered a method to classify ways that a human brain is organized to process modules of cognition.

Like SOAR, ACT-R relies on the implementation of a special coding language and access to an ACT-R interpreter. From here, one can produce an automatic simulation of human behavior, where cognitive operations like memory encoding, with visual and auditory stimuli as well as mental imagery manipulation, can be accounted for.

ACT-R’s declarative memory system has been designed to model human memory, and it can model the understanding and production of natural language. ACT-R has been used to capture how people solve algebraic equations

Applications of Cognitive Science and AI

There is a wide range of applications for cognitive science AI. The field has undergone a massive transformation over time, with resulting applications that are able to solve problems involving complex reasoning.

EvBrain is a mind simulation software designed to develop artificial brain models. Virtual animals can be created using this software and survive a predator-prey environment. An advanced level of intelligence is essential to processing large amounts of data and solving complex logic problems quickly and accurately. 

With the development of human-level intelligent agents, a replica of the human mind has been created that is easier to study than the human brain. The reason that an in-depth understanding of realistic simulations is needed, is to draw theories that showcase human nature and realize its cognitive limitations.

It’s hoped that the brain’s learning process will be better understood, including information retrieval, leading to improved learning methods. This knowledge would help with teaching and possibly with solutions for individuals that have suffered brain trauma. 

Speech and Text to Speech

The integration of cognitive science and AI started with speech to text services offering humans a diverse range of new capabilities. For years, systems of voice to text have helped with speech transcription, with the API recognizing the audio and providing real-time conversion to text. This has fostered smooth recording, but it now includes identification of the speaker, the time, and if any follow-up is required. 

Widespread industrial demand for text to speech and voice recognition systems have pushed forward Andriod’s native Google Assistant and One Note. The read-aloud feature provides users with audio rather than having to read the information in text form. The pitch, speech rates, and language are altered to suit the user. The Android Assistant and Apple’s Siri systems were created using neural network models, and there is a growing list of tasks they can perform. 


Another outcome of cognitive science in AI is personalized interactions that allow the AI to bring richer experiences when interacting with technology. More specifically, the AI is able to prioritize content that is constantly improving with more input from the user.

The more relevant and accurate the content that is provided, the more satisfaction that the user will have with the experience. Cognitive services, like those offered with Microsoft Azure, include Personalizer, which was created through reinforcement learning-based capability.

This service was inspired by a number of studies in the field of cognitive science, and reinforcement learning is a technique that allows AI to optimize goals based on individualized configurations. 

Simulating the Brain With AI

It is now possible for AI learning processes to become more intelligent with brain simulation models that mimic the human brain. However, investigating the functions of the brain using models is more complex.

The extreme precision, accuracy, and efficiency of millions of brain neurons are the reason for this complexity. Further, achieving this level of sophistication is difficult because most brain functions are analog transactions while computer models are entirely digital.

There are movements toward making analog computers, but this is far off for cognitive science needs.  This challenge means that simulating human brain becomes even more difficult, having to then generate algorithms that would be based on the working principles of the brain that are not the same as computers.

A second issue limiting computerized intelligence simulation is that the speed and capacity of hardware to perform the needed computations remains limited. No technologies can currently run large-scale simulations in real-time, but this may be improved with quantum computing. 

The Future of Brain Simulation AI

The next few decades will be an exciting time for AI and innovation. The human mind has the capacity to perform a vast array of different tasks that do not put significant stress on the brain. Nanotechnology that is aimed at increasing the memory and speed of computational hardware is expected to play a significant role in human brain simulation. Modern cognitive science and AI theories will be analyzed and ultimately foster a better understanding of the brain

Final Thoughts

AI is set to integrate itself completely into the human world. The goal of adding cognitive science into AI is to understand humans better. AI must think with empathy, and do so independently.

We hope it will develop this ability to learn new things and make decisions with emotional input and provide useful insights for humans. This is the true test for AI. 

Disclaimer: The author of this text, Jean Chalopin, is a global business leader with a background encompassing banking, biotech, and entertainment. Mr. Chalopin is Chairman of Deltec International Group,

The co-author of this text, Robin Trehan, has a bachelor’s degree in economics, a master’s in international business and finance, and an MBA in electronic business.  Mr. Trehan is a Senior VP at Deltec International Group,

The views, thoughts, and opinions expressed in this text are solely the views of the authors, and do not necessarily reflect those of Deltec International Group, its subsidiaries, and/or its employees.