“People are often surprised to learn the extent to which computer science is people-oriented”: Dr. Andrew McAllister on the Intersection of Computer Science and Psychology

(Photo provided by Andrew McAllister)

“The (computer science) field is not the study of computers: it’s all about using information to provide value for people’s lives. Computers and programming are just the tools we use to help make people’s lives more fun, convenient, profitable, or safer. A number of basic questions bring psychology and information technology together.” ~Andrew McAllister

Ψ

*In the following article, TS refers to The Synapse and AM refers to Andrew McAllister

TS: Could you briefly introduce yourself and what brought you to UNB?

AM: My name is Andrew McAllister. I am a professor in the Faculty of Computer Science at UNB. 

I grew up in Fredericton, so UNB was a natural choice for me after high school. I thought about becoming a psychotherapist, which is why my first degree is a BA in psychology. During my senior year I worked as a psychiatric attendant at the Chalmers hospital to gain some relevant experience. Unexpectedly, that experience convinced me that therapy wasn’t my first passion.

I had already taken a few introductory computer science courses as an undergraduate and I really enjoyed them. That led me to complete a Masters in computer science (CS) at UNB, during which I taught my first UNB CS course in 1983. (If you do the math, I’m rapidly approaching my 40-year anniversary. I still can’t quite believe that number!)

Those early teaching experiences went well enough that UNB offered me a faculty position if I would go away and complete a Ph.D. first. I did so … and here we are!

TS: What are your areas of expertise? 

AM: My psychology degree means I have an unusual background for a CS professor. I’ve always been drawn to topics near the intersection of the two areas. That includes a fair amount of work related to the early stages of software development, which relies on communicating with people to understand their needs - something called ‘systems analysis’. Specifically, it involves IT (information technology) workers getting together with non-IT folks who can provide expertise related to the type of software being developed. 

TS: How are the methods or topics of your area related or applicable to psychology?

AM: People are often surprised to learn the extent to which computer science is people-oriented. The field is not the study of computers: it’s all about using information to provide value for people’s lives. Computers and programming are just the tools we use to help make people’s lives more fun, convenient, profitable, or safer. 

It’s kind of like astronomers: they use telescopes as tools to help them study outer space … but astronomy is not the study of telescopes.

A number of basic questions bring psychology and information technology together. Here are just a few:

1. To what degree can we get computers to behave like people?

2. How can we design software so it best meets the psychological needs of users?

3. How does understanding the psychological needs of software developers help to make their work easier and more productive?

4. How can we enhance the capabilities of people using information technology?

A great deal of work has already been done in each of those areas. For instance:

1. Behaving like people: Artificial intelligence is helping us teach computers to learn on their own, make complex decisions, and other tasks that were once considered only for humans. Natural language processing (NLP) gives computers the ability to communicate with people in ways that are increasingly indistinguishable from talking with another person. For instance, have you had a conversation with your Alexa home device recently? Computers can also use cameras to see the world and interpret images, as humans can. This helps us scan barcodes and build self-driving cars, among hundreds of other applications.

2. Psychological needs of users: Much work has been done in the area of human-computer interaction (HCI). This helps us design screens, reports, voice simulations, virtual reality displays, gaming consoles, and seemingly endless other ways people can interact in natural ways with technology.

3. Psychological needs of developers: Software engineering is all about dealing with the inherently complex nature of computer code. Applications can involve thousands or millions of data points, decisions, and calculations. Machines can handle this just fine, but we humans can only keep a handful of thoughts in our mind at any one time. Much work has been done to bridge this gap.

4. Enhancing human capabilities: Examples of this have been developed in almost every area of human endeavour. Computer generated images (CGI) enhance the ability of artists to produce movie footage with a level of detail never seen before. Software helps fighter pilots fly jets that are otherwise inherently aerodynamically unstable. Artificial intelligence automates credit limit increase decisions for American Express when requested by credit card holders. 

TS: Which research topics in your area would you recommend to psychology students? Why?

AM: I spent a few days thinking about how I wanted to respond to these interview questions. During that time I happened to see an interview with Bill Gates, co-founder of Microsoft. He mentioned a few technologies that are likely to have a large impact on our future. The one he believes has the largest potential is artificial intelligence (AI).

That got my thoughts swirling: what sort of future AI development related to psychology would excite me?

I immediately thought of one of the most challenging areas of psychotherapy: couples therapy.

I read once that psychotherapy clients often provide overwhelmingly positive feedback about how the interventions helped them. This is not always the case, however, with marital therapists. 

There are many reasons for this. Couples often delay admitting their problems until things are so bad that dysfunction can’t be denied. Considerable pain tends to occur by the time they walk into a therapist’s office. The problems are much more difficult to solve than they would have been if help was sought earlier, but human nature and social stigma can get in the way.

Then, the couples face the considerable challenge of sharing accurate, complete, and unbiased information about their situation to the therapist. Unfortunately, incomplete memories - alongside emotions and opposing points of view - can make such reporting difficult.

Practitioners at the “Love Lab” in the Gottman Institute at the University of Washington try to bypass this. They gather information about clients by observing couples for several hours in a realistic apartment setting. However, this is still a limited snapshot in an artificial setting, and it is also beyond the means of most therapists.

Well … imagine a couple with a digital assistant in their home, similar to today’s Amazon Alexa or “Hey Google”. Such a setup could capture relationship interactions over long periods of time. This could perhaps be enhanced with video capture to read body language and facial expressions.

Artificial intelligence techniques could be used to interpret interactions. This could help identify early warning signs of relationship trouble well before they fester and become larger issues. This could also help provide a more accurate view into how the couple actually interacts, including detecting moments when unnoticed slights and hurts enter the picture.

Such a system could report information that couples could use to thwart trouble before it snowballs. Or it could produce a report they could take to a marital therapist, providing a more complete and unbiased history of their interactions.

Oh, what the heck – let’s go there: how about a digital therapist? We could load up the AI with the current best practices in marital interventions. Dr. Sue Johnson’s work on Emotionally Focused Couple Therapy (EFT) comes to mind as one possible example, and there are undoubtedly others.

Imagine one partner making a particularly egregious relationship gaffe, then the “Relationship GPS” lights up and says: “Recalculating …”

Is this starting to sound frivolous? Or too creepy or unreliable to imagine anyone using such technology? Are there significant privacy issues?

Of course, those are serious concerns that would need to be resolved, but this is the type of research vision that technology is rapidly bringing into the realm of the possible. Just as human behaviour is infinitely variable, there are countless research opportunities at the boundary between psychology and computer science. Not all of them are as far “out there” as the example I conjured up.

Participating in this type of work does not require a psychologist who also has a deep understanding of IT. I can imagine a multi-disciplinary team of computer scientists, psychologists or therapists, ethicists, and maybe even lawyers coming together to work on this type of project. Each would bring their own skills and knowledge to the table.

TS: Which courses in your area would you recommend to psychology students? 

AM: Psychology students interested in understanding a bit more about IT might consider starting with our overview of computer science – CS1203. Students interested in developing technical skills can begin with our introductory first-year Java programming courses – CS1073 and CS1083.

Senior courses directly related to psychology include:

1. CS3025 – Human-Computer Interaction

2. CS4725 – Introduction to Artificial Intelligence

3. CS4765 – Natural Language Processing

Warning – these senior courses require quite an investment in terms of taking prerequisite CS courses.

TS: Is there anything else you wish to share about how your area overlaps with psychology?

AM: How much does CS overlap with psychology? In my opinion, the answer is: way more than people might think! Ψ

Created for The Synapse by Incé Husain.