Artificial intelligence (AI) has improved significantly, from powering self-driving cars to assisting with medical diagnosis. However, an important question remains: Can AI pass cognitive tests designed for humans? Despite impressive results in areas such as language processing and problem solving, AI still strives to replicate the complexity of human thinking.
AI models like chatgpt can generate text and solve problems effectively, but they perform poorly when faced with cognitive testing such as Montreal Cognitive Assessment (MOCA) (designed to measure human intelligence).
The gap between AI’s technological achievements and cognitive limitations highlights the major challenges of its potential. AI has not yet matched human thinking, especially in tasks that require abstract reasoning, emotional understanding, and contextual awareness.
Understand cognitive testing and its role in AI evaluation
Cognitive testing (such as MOCA) is crucial to measuring all aspects of human intelligence, including memory, reasoning, problem solving, and spatial awareness. These tests are often used in clinical settings to diagnose conditions such as Alzheimer’s and dementia, leading to a deeper understanding of how the brain functions in different situations. Tasks such as recalling words, drawing clocks, and recognizing patterns evaluate the brain’s ability to complex environments that are essential in everyday life.
However, when applied to AI, the results are very different. AI models (such as Chatgpt or Google’s Gemini) may do well in tasks such as identifying patterns and generating text, but they struggle with cognitions that require a deeper understanding. For example, while AI can follow clear instructions to accomplish tasks, it lacks the ability to abstract reasoning, explain emotions, or apply context, which is a core element of human thinking.
Therefore, cognitive testing has a dual purpose in evaluating AI. On the one hand, they emphasize the advantages of AI in processing data and effectively solving structured problems. On the other hand, they have a big gap in the ability of AI to replicate human cognitive functions, especially those involving complex decision-making, emotional intelligence, and contextual awareness.
With the widespread use of AI, its application in areas such as healthcare and autonomous systems requires not only the completion of tasks. Cognitive testing provides a benchmark for evaluating whether AI can handle tasks requiring abstract reasoning and emotional understanding, which is the quality of the human intelligence center. For example, in healthcare, while AI can analyze medical data and predict disease, it cannot provide emotional support or make subtle decisions that depend on understanding the patient’s unique situation. Similarly, in autonomous systems such as autonomous vehicles, explaining unpredictable scenarios often requires human intuition, while current AI models lack it.
Using cognitive tests designed for humans, researchers can identify areas where AI needs to improve and develop more advanced systems. These assessments also contribute to realistic expectations about what AI can do and emphasize that human participation remains vital.
AI limitations in cognitive testing
AI models have made impressive progress in data processing and pattern recognition. However, these models face significant limitations in tasks requiring abstract reasoning, spatial awareness, and emotional understanding. A recent study that used Montreal Cognitive Assessment (MOCA) to test multiple AI systems, a tool designed to measure human cognitive abilities, revealed a clear gap between AI’s strengths in structured tasks and the struggle with more complex cognitive functions.
In this study, Chatgpt 4o scored 26 out of 30 points, indicating mild cognitive impairment, while Google’s Gemini scored only 30 points, reflecting severe cognitive impairment. One of the most important challenges for AI is visual-spatial tasks such as drawing clocks or copying geometric shapes. These tasks require understanding spatial relationships and organizing visual information, and are the areas in which humans express intuitively. Despite clear instructions, AI models are still working to complete these tasks accurately.
Human cognition integrates sensory input, memory, and emotions, so that decision-making can be adaptive. People rely on intuition, creativity and environment when solving problems, especially in ambiguous situations. This ability to think abstractly and use emotional intelligence in decision-making is a key feature of human cognition, thus enabling individuals to navigate complex and dynamic scenarios.
Instead, AI works by processing data through algorithms and statistical patterns. Although it can generate responses based on learning patterns, it does not really understand the context or meaning behind the data. Lack of understanding makes it difficult for AI to perform tasks that require abstract thinking or emotional understanding, which is crucial in tasks such as cognitive testing.
Interestingly, cognitive limitations observed in AI models are similar to disorders in neurodegenerative diseases like Alzheimer’s. In this study, when asked about spatial consciousness, AI was overly simple and context-dependent, similar to individuals with decreased cognitive abilities. These findings emphasize that while AI is good at processing structured data and making predictions, it lacks the depth of understanding required for more nuanced decisions. Such limitations particularly involve health care and autonomous systems where judgment and reasoning are crucial.
Despite these limitations, there is potential for improvement. Newer versions of AI models, such as ChatGpt 4O, have shown progress in inference and decision-making tasks. However, replicating human-like cognition will require improved AI design, possibly through quantum computing or more advanced neural networks.
Artificial intelligence struggles with complex cognitive functions
Despite advances in AI technology, it still has a long way to go from passing cognitive testing designed for humans. Although AI is good at solving structured problems, it lacks for more nuanced cognitive functions.
For example, AI models often miss marks when asked to draw geometric shapes or interpret spatial data. Humans naturally understand and organize visual information, and AI strives to do this effectively. This highlights a fundamental problem: AI’s ability to process data does not mean understanding how human thinking works.
The core of AI limitations is based on the nature of algorithms. AI models operate by identifying patterns in the data, but they lack the contextual awareness and emotional intelligence that humans use to make decisions. Although AI may effectively generate outputs based on what is trained, it does not understand the meaning of the output behind these outputs. This inability to abstract thinking, coupled with a lack of empathy, can prevent AI from completing tasks that require deeper cognitive functions.
This gap between AI and human cognition is evident in healthcare. AI can assist in tasks such as analyzing medical scans or predicting diseases. Nevertheless, it cannot replace human judgment in complex decisions, involving understanding the patient’s situation. Similarly, in systems such as autonomous cars, AI can process large amounts of data to detect obstacles. Nevertheless, when making split decisions in unexpected situations, it cannot replicate the intuition that humans rely on.
Despite these challenges, AI still shows potential for improvement. Newer AI models begin to deal with more advanced tasks involving reasoning and fundamental decision-making. But even if these models advance, they remain inconsistent with the wide range of human cognitive abilities required to pass cognitive tests designed for humans.
Bottom line
In short, AI has made impressive progress in many areas, but there is still a long way to go before passing cognitive testing designed for humans. Although it can handle tasks like data processing and problem solving, AI struggles with tasks that require abstract thinking, empathy, and contextual understanding.
Despite improvements, AI is still struggling with tasks such as spatial awareness and decision-making. Although AI has hope for the future, especially with technological advancement, it is far from replicating human cognition.
