By Atefeh Ferdosipour
When considering conceptual frameworks in cognitive psychology that connect to technology, it’s clear that information processing theories are the most relevant.
Artificial intelligence, at its root, is essentially the same as the computers that were introduced to the market several decades ago, with programming to make conditions easier for learning and various types of human activities. These machines, which included heavy monitors and keyboards, were made of hardware and software components. And of course, today they are also available in foldable, very light, but much more advanced forms, in various models.
But if you are curious about the intellectual origin and the design model of these computers, the answer is always the same: The structure of the human mind was the model for their design.
More specifically, there is a perspective in psychology and learning sciences called “information processing” that is the real origin of the idea of building computers. And if we consider computers as the unquestionable parent of the idea of artificial intelligence—which they really are—then the closest conceptual framework to artificial intelligence is this same information processing theory.
Before going into the importance and application of information processing theory in artificial intelligence and generating practical ideas for designing AI tools, this article will explain the conceptual framework of this perspective. This review will make clear why information processing is the most relevant theory of cognitive and learning sciences to technology.
The Place of Information Processing Theory in Learning Sciences
Information Processing Theory is an interdisciplinary bridge that connects cognitive psychology, computer science, linguistics, and some other fields, which emerged in the 1950s–1960s. Like other cognitive perspectives, which believe that to understand deep learning in humans and important concepts such as problem solving, thinking, and the complexities of higher-level learning, we should use cognitive principles rather than behavioral ones, this perspective also emphasizes the study of the human mind.
Among several theorists of this perspective, a common point has been clear: The human mind is an active information processing system, not a passive receiver!
In other words, the learner has an active system for attention, reception, organization, processing, and storage of information. That is why a combination of raw data and input, plus additional received information, can produce a different output and final product!
If we want to find the best analogy for this kind of functioning, we might return to a cognitive perspective and conceptual framework that inspired the construction of early computers.
In other words, just as a computer has software and hardware components, the mind is also a complex box that takes input information and, with an active program and supported by hardware (the brain and nervous system), works on it; the results are understanding a particular topic, generating new ideas, and other high-level mental activities.
The Idea & Structure of Human Information Processing
The general framework of this theory is similar to early computer models and telecommunication systems.
According to this framework, the human mind, like a telecommunication system or a computer, receives and interprets messages, encodes them, stores them, and reconstructs them. In humans, the “messages” or information, is received through the senses.
The central idea is this: What distinguishes humans from machines is that machines are passive receivers of information, while humans are active processors of information.
The main stages or components of this structure are described below.
1. Sensory Memory
The first stage in information processing is sensory memory. In this stage, raw information obtained through the senses (vision, hearing, touch, etc.) is held in the system for a very short time—fractions of a second to a few seconds.
The capacity of this memory is unlimited, but its duration is a fraction of a second. Anything may be received, but if it is not attended to, it quickly disappears.
2. Attention: The Gateway to Deeper Processing
Attention is the mechanism that determines which sensory information enters the later stages of processing. In other words, attention is the gateway into the information processing system.
A very important point in this stage is that factors such as individual goals and expectations, motivation and emotional states, and the context and situational background at the time of receiving information are among the most important determinants that cause the user or learner to direct sensory information toward deeper processing.
So the necessary condition for meaningful processing is attention, and the prerequisite for attention is the factors mentioned above.
3. Working or ‘Short-Term’ Memory
Working memory, or short-term memory, is a bridge between attention to information and preparing information to enter the stage of deep processing in the long-term store.
Although the capacity of this store is short, it plays a strategic part for preparing information retention.
Information is filtered in this stage; if it exceeds capacity, it is chunked to make room for more information. Proponents of this perspective believe that information obtained through visual or auditory coding dominates other information (for example, information that has entered through visual or acoustic cues).
Although working memory is limited and lasts no more than a few seconds to a few minutes, it is very strategic in terms of filtering information; that is why it is called “active” working memory.
4. Encoding: The Bridge Between Working and Long-Term Memory
In order for information to move from short-term to long-term memory, it must undergo deep processing. The condition for this is attention to the meaning of the information, not only to superficial and/or detailed aspects of the content.
Deep encoding of information happens through attention to meanings, connecting new information with old, organizing and structuring information, and, finally, classifying information into semantic categories.
For this reason, followers of this perspective believe that techniques such as using examples, using concept maps and diagrams, multiple representations, and any method that gives information an organized semantic structure can lead to deep processing and long-lasting learning.
5. Long-Term Memory
Long-term memory is a relatively stable store of knowledge, skills, and individual experiences. Its duration can range from a few minutes to a lifetime.
This memory has no limitation in terms of duration or capacity!
Types of long-term memory:
- Declarative memory: Knowledge about “what”
- Semantic memory: General knowledge, concepts, facts
- Episodic memory: Personal experiences and specific situations
- Procedural memory: Knowledge about “how” to do things (skills, habits)
This memory is the foundation of mental structures and even affects future learning. Information in long-term memory is stored as stable and structured schemas that interpret and make sense of later information, inputs, and experiences of the learner. So this store determines how new information is understood, interpreted, and encoded.
Accordingly, the process of deep learning means reorganizing this part of memory.
6. Retrieval: When Learning Appears in Practice
Learning is considered “real” when information stored in long-term memory can be retrieved and used in appropriate situations. Retrieval is not merely remembering; it includes using knowledge in problem solving, decision making, reasoning, and action.
Types of retrieval:
- Pure recall
- Applying old information in a new situation
- Reconstructing and changing the structure of information
The more precise the encoding in earlier stages, and the more meaningfully the information units are structured, the faster and easier retrieval will be in this stage.
7. Control Processes & Cognitive Strategies
More advanced models of the information processing perspective believe that, in addition to the components mentioned earlier, humans are equipped with a set of control processes and cognitive strategies.
These include planning, monitoring information, changing strategies, and evaluation.
These processes are closely related to concepts such as metacognition and self-regulation in learning. These concepts are outside the scope of this article.
Deep Learning in the Framework of Information Processing Theory
As mentioned earlier, learning includes stages and processes, from receiving and attending to information to durable storage and memorization.
But deep learning, within the framework of information processing theory, is a chain of activities that causes new information to be stored deeply and in the form of semantic structures, and for this information to remain usable and retrievable for a long time.
This deep and stable learning occurs when:
- Attention to input information (raw materials) is properly directed
- Pressure on working memory is in line with its capacity
- Semantic codes are used more than short-term auditory and visual codes
- Information is transformed from a disorganized state to a structured and coherent state
- That stored information is used and applied in various situations
Strengthening strategic skills for monitoring information will also accelerate this process.
Applying Principles from the Information Processing Perspective to Designing Learning Systems and AI Tools
To use the information processing perspective in learning design or for designing chatbots and other artificial intelligence tools—which are not limited to educational environments and are applicable everywhere—we must focus on the principles mentioned above and the human information processing mode and try to extract the most relevant practical principles from them.
In this section, I have formulated and suggested some of the most applicable principles in this area for designers.
From an information processing perspective, the design of AI models must consider the active interaction between human and machine. It’s as if the artificial intelligence acts as an assistant to the process of information processing in the learner’s (or user’s) mind.
Managing and designing this interaction must be done in the following areas, and, according to the natural model of human information processing, it must be human-centered:
1. Managing the User’s Attention Flow
In order to attract maximum attention, the design should include clear signals for guiding and capturing attention. Obviously, using mechanisms such as posing diverse and curiosity-provoking questions, giving short but attractive hooks, avoiding giving immediate and fast answers, and providing small challenges for the user can be effective in this direction.
Also, from the perspective of information processing theory, the issue is not how much information is given, but how much of that information is actually processed.
Among other methods for proper use of the user’s time and attention, the designer should define clear learning goals for the user and highlight important points with cues and key questions. Avoiding information overload and distractions like side notes and explanations, increases focus.
2. Considering the Limitations of Working Memory
Given the limited capacity of short-term or working memory, it is better to avoid cognitive and information overload. Several key strategies are suggested:
- Breaking and chunking information into smaller sections and smaller steps
- Breaking long text up with subheadings
- Using structured information instead of a mass of detail
- Helping the user to receive and attend to information in steps and staged paths
3. Strengthening Encoding: Creating Opportunities for Meaningful Processing
Artificial intelligence, instead of just giving answers, can act as an encoding facilitator.
Among the effective strategies suggested in this direction are:
Using guiding questions
- “Can you summarize this idea in your own words?”
- “What similarity does this concept have with something you already knew?”
- “Which part was newer for you? Why?”
Ask users to explain how to apply the idea in the real world, e.g., “Give an example from your own work or study situation where this idea could be applied.”
Ask users to summarize and restate the information in their own language, for example, in two or three lines.
4. Organizing Knowledge and Information
The organizing process means helping the learner to build coherent frameworks and mental maps.
As a few effective suggestions in this direction, we can use the following strategies:
- Use artificial intelligence to provide retrieval practice: This is fully achievable by including questions, challenges, and situations that force the learner to pull information out of memory.
- Use artificial intelligence tools to provide meaningful feedback. For this purpose, feedback that corrects not only the result but also the strategy and learning process is more meaningful and effective.
- Use artificial intelligence to teach learners self-regulation strategies and information evaluation alongside the main content; for example, telling them: “Choose the different memorization methods that are more effective for you, and finally evaluate yourself in this process and select the method that suits you.”
- Ask users to categorize and organize the information in their own way.
Such interactions can cause information to move from superficial and short-term processing toward deep and durable processing.
5. Designing Smart Feedback
Within the information processing framework, feedback plays a vital role in regulating the processing and learning process. Such feedback, both in facilitating the process and in motivating the user to develop cognitive skills and self-evaluation, is highly effective.
Accordingly, it is suggested that artificial intelligence can design feedback that:
- Not only points out mistakes in understanding and comprehension, but also reminds the user of the reasons for their mistakes. This type of feedback is much more effective than simple “right/wrong” feedback.
- In addition to pointing out mistakes, AI-generated feedback might suggest corrective strategies and use the user’s past and successful experiences to select them. For example, it can ask: “In such situations, what techniques have you used successfully?”
- Instead of emphasizing the final product and learning outcome, artificial intelligence might emphasize the processes and strategies of receiving information.
- Feedback should be based on performance rather than the final product and final learning. To accomplish this, use more strategic questions and focus on methods and appropriate feedback for these—rather than questions and answers about content!
By designing more effective feedback, learning supports self-regulation and the users’ development of cognitive and metacognitive skills. Learners become more active and self-initiated in the process of interacting with artificial intelligence and return from the state of “mere receiver” to their natural and real nature as an “information processor.”
And finally, in addition to acquiring new knowledge and information about a particular topic, users also learn new mental techniques and strategies, or their previously successful strategies are strengthened.
Conclusion
My previous articles focus on the application of experience-based theoretical perspectives in AI design; I believe that we can use the effective and useful aspects of theories, even old ones, in the age of technology.
This article applies information processing theory to AI-supported learning design. I believe that human-centered artificial intelligence is precisely about this: When the information processing perspective aligns with the natural process of human information processing, technology becomes more efficient and more responsible.
The ultimate goal is for information delivery to become information processing design—a constructive bridge for moving closer to human-centered artificial intelligence.
Based on the information processing perspective, artificial intelligence can:
- Direct attention
- Manage pressure on working memory
- Create opportunities for meaningful encoding
- Facilitate retrieval and application
- Support self-regulation and metacognition
If we apply these principles in designing tools and learning experiences based on artificial intelligence, we can use AI not only as a content production engine, but also as a partner and learning designer that is aligned with how the human mind works.
References
Ashcraft, M. H., & Radvansky, G. A. (2013). Cognition (6th ed.). Pearson.
Baddeley, A., Eysenck, M. W., & Anderson, M. C. (2020). Memory (4th ed.). Routledge.
Eysenck, M. W., & Keane, M. T. (2020). Cognitive Psychology: A Student’s Handbook (8th ed.). Routledge.
Ferdosipour, A. (2026). Can AI Guide Deep Learning? An Answer From Gestalt Cognitive Psychologists. The Learning Guild.
Ferdosipour, A. (2026). Why AI Needs Vygotsky: The Case for AI-Based Intentional Friction. The Learning Guild.
Gluck, M. A., Mercado, E., & Myers, C. E. (2020). Learning and Memory: From Brain to Behavior (4th ed.). Worth Publishers.
Learn to Create More Effective Learning at the Evidence-Informed Design Online Conference
Don’t miss the Evidence-Informed Design online conference, October 14–15, where keynote Nidhi Sachdeva will explore cognitive architecture as a foundation for effective design. Across six sessions, you’ll learn why and how to reduce extraneous load, write clearer narration, build reflection and practice into learning, and use feedback effectively. Register now; online conferences are free for Professional or Enterprise members!
Image credit: Nattapon Kongbunmee
