By now, you’ve probably been convinced that AI (artificial intelligence) provides all the opportunities one can hope for and all the threats one should fear. But with all the news about AI coming to us, it almost seems like AI is a recent invention that suddenly took us by surprise and will change society forever. This is an independent discipline recently introduced. I do not deny the possibilities offered by AI, as I would not deny the promise that no other scientific discipline or method offers. And I’m certainly not ignorant of the risks that AI poses to us, as I wouldn’t ignore any of the risks that no technology poses to us.
The history of AI
Source: From Pixabay’s Wirestock
Artificial intelligence and cognitive science are very interconnected. AI was first introduced in a proposal for a seminar held in the summer of 1956 at Dartmouth College. The workshop aimed to find answers to questions such as how to make machines use language, form abstractions and concepts, and solve problems: How artificial minds that work like human minds can be developed. The meeting was attended by 11 computer scientists, including Alan Newell and Herb Simon, who won the Nobel Prize for their work two decades later.
A special interest group in information theory was held a few weeks after the Dartmouth seminar at the Massachusetts Institute of Technology. The meeting was attended by researchers in psychology, linguistics, computer science, anthropology, neuroscience and philosophy. At this meeting, questions about language, abstractions and concepts and problem solving played a central role. These questions were motivated by the central question: How can we better understand the human mind by developing artificial minds. The meeting of the special interest group was attended by several researchers who also attended the seminar in Dartmouth.
The MIT Task Force marks the cognitive revolution and the beginning of cognitive science. The cognitive revolution can best be characterized by a movement that emphasizes the interdisciplinary study of the human mind and its processes, with special emphasis on the similarities between computational processes and cognitive processes, between human minds and artificial minds. This led to what is now known as cognitive science, an interdisciplinary research program consisting of psychology, computer science, neuroscience, linguistics and related disciplines.
Looking back on these encounters in the 1950s, the birth of artificial intelligence and the birth of cognitive science, it seems almost as if AI is a computer science motivated by psychology and the psychology of cognitive sciences motivated by computer science.
Concepts and methods
The connection between AI and cognitive science is not limited to two seminars. There are also striking similarities in the theories, concepts, and methods they both use.
Reinforced learning in AI obviously stems from reinforced learning as we know it in psychology. And central to AI today is deep learning, the use of artificial neural networks. These artificial neural networks are inspired by human neural networks. Especially around the 1980s, these artificial neural networks showed a lot of promise, less at that time in AI and more in cognitive science.
Cognitive Science / AI
Source: From Pixabay’s GDJ
The connection between AI and cognitive science can also be seen against the background of leading researchers. Among the researchers who proposed the Dartmouth seminar were John McCarthy, both a computer scientist and a cognitive scientist, and Marvin Minsky, a both cognitive and computer scientist. Others who attended the seminar, including Alan Newell, had experience in psychology and computer science. David Rumelhart and Jay McClelland, who led the study of artificial neural networks in the 1980s, both have experience in psychology. And one of the contributors to Rumelhart and Jay McClelland’s two-volume Parallel Distributed Processing was Jeff Hinton, considered one of the leading figures in artificial neural networks, a cognitive psychologist, and a computer scientist.
The price of explanation
But there is a more important message about the interdependence of AI and cognitive science. This message does not lie in the history of AI and cognitive science, nor in the use of such concepts and methods, nor in the background of researchers. It lies in what we can learn from AI and cognitive science. For example, regarding the importance of Explainable AI, also called XAI. While AI (and data science) often focus on accuracy, we may want to pay more attention to why techniques and methods make specific decisions.
The accuracy and performance of AI systems are rapidly increasing with more computing power and more sophisticated algorithms. This also has a price: clarity. If we want to build algorithms that are fair – following the FAIR principles of discovery, accessibility, interoperability and reuse of digital assets – we must at least be able to understand the mechanisms behind the algorithms. And it reminds me of what Mike Jones said a few years ago about data science:
In the cognitive sciences, we have been significantly more skeptical about the promise of big data, mostly because we attach so much value to explanation to prediction. The main goal of every cognitive scientist is to fully understand the system under study, instead of being satisfied with a simple descriptive or predictive theory. (Jones, 2017)
In other words, the interdependencies between AI and cognitive science are not just in the past. In fact, more than ever, they lie in the present and the future.