Site navigation

Glasgow University Study Helps Develop AI that “Thinks” Like Humans

Michael Behr


Glasgow University AI
By comparing how AIs and people process and understand visual information, scientists can create systems that are more accurate and make fewer unexplained mistakes.

Recent research from Glasgow University could help artificial intelligence (AI) move towards mimicking human behaviour and thought processes.

The new study saw a research team use 3D modelling to analyse how Deep Neural Networks – a form of machine learning – process information. This helped the team visualise the similarities between how people and AIs process information.

Deep Neural Network technologies are frequently used in applications such a facial recognition. However, despite their success, scientists still do not fully understand how these networks process information, and therefore errors may occur.

Researchers modelled the visual stimulus received by the Deep Neural Network and transform it in multiple ways. This helped them demonstrate a similarity of recognition, via processing similar information, between humans and the AI model.

The researchers asked a group of people to compare how similar a series of 3D faces were to four familiar identities. The Deep Neural Networks then made judgements, and the results were compared to the humans’ decisions.

This allowed the researchers to determine whether the machines made the same ratings as the people for the same reasons – testing not only whether humans and AI made the same decisions, but also whether they were based on the same information.

Senior author and head of the University of Glasgow’s Institute of Neuroscience and Technology Professor Philippe Schyns said: “When building AI models that behave “like” humans, for instance to recognise a person’s face whenever they see it as a human would do, we have to make sure that the AI model uses the same information from the face as another human would do to recognise it.

“If the AI doesn’t do this, we could have the illusion that the system works just like humans do, but then find it gets things wrong in some new or untested circumstances.”


Understanding how artificial intelligences process information is one of the key challenges facing AI development. Humans and machines “think” in different ways, making it difficult to apply models and metaphors from one field to the other.

As such, being able to visualise and compare similarities and differences between how the two process information can help researchers develop more human-like AIs, capable of performing a wider range of tasks.

For example, while Deep Neural Networks can achieve or exceed human performance in some tasks, they fall behind in comparatively simple visual discrimination tasks when compared to humans.

The researchers hope their work will help create more dependable AI technology that will process information like humans and make errors that people can understand and predict.

The study, led by the University of Glasgow’s School of Psychology and Neuroscience, was published in the journal Patterns.

Get the latest news from DIGIT direct to your inbox

Our newsletter covers the latest technology and IT news from Scotland and beyond, as well as in-depth features and exclusive interviews with leading figures and rising stars.

We will keep you up to date on the pivotal issues impacting the sector and let you know about key upcoming events to ensure that you don’t miss out on what’s going on across the Scottish tech community.

Click here to subscribe.

Michael Behr

Senior Staff Writer

Latest News

%d bloggers like this: