Many people think of psychology as primarily thinking about mental health, but that story goes far beyond that.
As a science of the mind, psychology played a pivotal role in shaping artificial intelligence, providing insights into human cognition, learning, and behavior that have had a major impact on AI development.
These contributions not only lay the foundations of AI, they continue to guide future developments. Psychological research shapes an understanding of what constitutes machine intelligence and how it can address the complex challenges and benefits associated with this technology.
Machines that mimic nature
The origins of modern AI can be traced back to psychology in the mid-20th century. In 1949, psychologist Donald Heb proposed a model of how the brain learns.
This idea gave us some hints on how machines learn by mimicking the natural approach.
In the 1950s, psychologist Frank Rosenblatt developed a system called the Perceptron based on Heb’s theory.
Perceptron was the first artificial neural network ever created. It was performed on the same principles as modern AI systems where computers learn by adjusting connections within the network based on data rather than relying on programmed instructions.
Scientific understanding of intelligence
In the 1980s, psychologist David Rumelhart improved Rosenblatt’s perceptron. He applied a method called backpropagation. It uses computational principles to help improve neural networks through feedback.
Backpropagation was originally developed by Paul Werbos. Paul Verbos said the technique “discovers the possibility of a scientific understanding of intelligence that it is important for psychology and neurophysiology, just as Newton’s concept was for physics.” .
Rumelhart’s 1986 paper, co-authored by Ronald Williams and Jeffrey Hinton, is often attributed to spark the modern era of artificial neural networks. This work laid the foundation for deep learning innovations such as large-scale language models.
In 2024, the Nobel Prize in Physics was awarded to Hinton and John Hopfield for their work on artificial neural networks. In particular, the Nobel Committee highlighted the important role that psychologists played in the development of artificial neural networks in the scientific report.
Hinton, who holds a degree in psychology, admitted to standing on the shoulders of giants such as Rumelhart when he was awarded.
Self-reflection and understanding
Psychology continues to play an important role in shaping the future of AI. It provides theoretical insights to address some of the biggest challenges in the field, including reflexive reasoning, intelligence, and decision-making.
Microsoft founder Bill Gates recently pointed out some important limitations on today’s AI systems. They cannot engage in reflexive reasoning, or what psychologists call metacognition.
In the 1970s, developmental psychologist John Fravel introduced metacognitive ideas. He used it to explain how children can acquire complex skills by reflecting and understanding their thoughts.
Decades later, this psychological framework has attracted attention as a potential pathway to advance AI.
Fluid Intelligence
Psychological theory is increasingly being applied to improve AI systems, particularly by increasing the ability to solve new problems.
For example, computer scientist François Charette emphasizes the importance of fluid intelligence, which psychologists define as the ability to solve new problems without prior experience or training.
In a 2019 paper, Chollet introduced tests inspired by the principles of cognitive psychology to measure how well AI systems can handle new problems. The test, known as The Abstract and Reasoning Corpus for Artificial General Intelligence (ARC-AGI), provided a kind of guide to making AI systems think and reason in a more human-like way.
In late 2024, Openai’s O3 model showed significant success in Chollet’s testing, showing progress in the creation of AI systems that can adapt and solve a wider range of problems.
Risk of explanation
Another goal of the current research is to enable AI systems to explain more output. Again, psychology offers valuable insights.
Computer scientist Edward Lee draws on the work of psychologist Daniel Kahneman and highlights why it can be dangerous for AI systems to require that they explain themselves.
Kahnemann showed how humans often justify their decisions with explanations created after the fact. For example, studies have found that judges’ rulings vary depending on the last meal despite their firm belief in their own fairness.
Lee warns that AI systems can produce similarly misleading explanations. Because rationalization can be deceptive, Lee argues that AI research should instead focus on reliable outcomes.
Techniques that shape our minds
The science of psychology remains widely misunderstood. For example, in 2020, the Australian government proposed to reclassify it as part of the university’s humanities.
As people interact more and more with machines, AI, psychology and neuroscience may hold important insights into our future.
Our brains are highly adaptable and technology shapes the way we think and learn. For example, a study by psychologist and neuroscientist Eleanor Maguire reveals that the brains of taxi drivers in London are physically altered by using cars to navigate complex cities. Ta.
As AI progresses, future psychological research may reveal how AI systems can enhance our capabilities and unleash new ways of thinking.
Recognizing the role of psychology in AI can promote a future in which people and technology work together for a better world.
This article will be republished from the conversation under a Creative Commons license. Please read the original article.
