A short circuit, such as the one shown, results in Johnny Five's artificial intelligence in the film (image by MdE via Flickr)

Science at the Movies: Short Circuit

Discussing artificial intelligence with U of T's experts

Before there was Wall-E, before Star Trek's Commander Data, there was Johnny Five: the sentient robot starring alongside Ally Sheedy and Steve Gutenberg in Short Circuit, a family-friendly film from 1986 that introduced a generation of kids to the wonders of artificial intelligence.

Tomorrow night, Johnny Five will drive discussion between audience members and U of T experts at the latest instalment of Science at the Movies.

It's the third and final in a series of screenings organized by Professor Ray Jayawardhana, Canada Research Chair in Observational Astrophysics and Senior Advisor on Science Engagement to U of T President David Naylor.

Each screening includes a question and answer session moderated by science journalist Dan Falk.

Philosophy professor Diana Raffman and computer science professor Sheila McIlraith will help spark the discussion about artificial intelligence (AI) with Falk and the audience.

McIlraith told U of T News about what she hopes to bring to the debate tomorrow night.

What interests you about Short Circuit?
The movie interests me for several reasons. From a philosophical perspective, it raises a number of interesting issues relating to how we define intelligence, how we determine an artifact to be intelligent, and whether an artificial intelligence can attain consciousness. These are fascinating issues that have been discussed and debated for a number of years, and popularized through the Turing test, Searle’s Chinese Room, and other works. The movie also interests me from a technological perspective. It was released in 1986. We’ve made significant advances in AI and robotics since that time. It was fun to see the robots through the eyes of the movie developers. People often think that what is hard for us to do as humans will be hard to get a robot to do, and more often than not it’s the opposite. You see this perspective reflected in the movie’s robots. Some of the things they could do seamlessly remain impossible today, whereas other things that were hard for the robots are now trivial.

How have discussions around artificial intelligence evolved since Short Circuit?
We are perhaps less concerned about the philosophical issues associated with whether an AI can attain consciousness and more focused on “just doing it”. The philosophical issues remain of interest, but are less at the fore. On the other hand, as we’ve had greater success, there has been renewed interest and concern about ensuring that people understand the capabilities and limitations of the technology being developed. Indeed, in 2008, then President of the Association for the Advancement of Artificial Intelligence (AAAI) struck a presidential panel to discuss so-called long-term AI futures.

How effectively does the film raise questions of consciousness and artificial intelligence?
I think it does a good job of raising questions of consciousness and AI. As an example, in the movie, robot Number 5 (later known as Johnny Five) proclaims itself to be “alive.” Its creator, scientist Newton Crosby explains quite rationally why this cannot be the case, but then puts Number 5 through a form of Turing test, to assess Number 5’s ability to “think.”  

How do your own research interests relate to themes raised in the movie?
My research is concerned with how we can create artificial intelligences with higher level cognitive functions. The broad area is often referred to as cognitive robotics. I saw a number of instances of the types of problems that we work on. For example, at one point Number 5 had to diagnose what was wrong with itself and actually rewired its circuitry. The task of automated diagnosis, self-reconfiguration, and self-repair has been a theme of some of my research. Similarly, Number 5 demonstrated human-like planning and decision-making capabilities as it successfully evaded the unlikable Captain Skroeder and his troops. This also relates to issues my research group addresses, though our planning and decision-making tasks are more benign.

Any other favourite movies dealing 'intelligently' with AI that you'd recommend? 
There are so many! 2001:  A Space Odyssey is a classic, as is Blade Runner. I also really liked Terminator 2, and I can’t leave out Star Trek, though I preferred the TV series to the movies.

The Bulletin Brief logo

Subscribe to The Bulletin Brief