Humans have a lot to think about and even more to process when it comes to solving the problems we face. Somewhere in all the data we churn out is everything from a HIV cure to the perfect videogame character. Computers with narrowly-programmed parameters have the speed but not the creativity to find what we need amid messy abstracts, so the theory behind AI is to give machines human-like powers of extrapolation, a ‘what if’ quality to consider the not-always-obvious. Such ‘short cutting’ of data processing will fundamentally change human decision-making.
Careers in AI are available in both the research/academic sectors and industry, where large IT firms like Google and Microsoft are funding exciting research. So far the work being done is mostly concerned with improving the software algorithms on which AI is based or interfacing it with hardware more effectively, such as in robotics.
Because AI is going to impact so many areas, you can study it in a wide variety of majors or degrees. The University of Western Australia’s (uwa.edu.au) AI unit, for example, is part of both the Computation and Entertainment Technologies majors.
Curtin University of Technology’s (curtin.edu.au) course offering is typical of the part AI has to play in both hardware and software. Their Bachelor of Engineering (Computer Systems Engineering) is a more practical course where you’ll build AI theory into machinery that executes it, whereas the Bachelor of Science (Computer Science) exposes students more to the algorithm analysis and software design that underpins AI efforts.
Edith Cowan University’s (ecu.edu.au) School of Computer And Information Science offers an introductory AI unit that will arm you with an overview of the field and help you decide where to specialise whether it’s theory, software, robotics or data mining.
The Monash University Information Technology faculty offers a comprehensive AI unit that includes everything from AI history and philosophy to programming milestones like fuzzy logic, machine learning and evolutionary algorithms. A more practical approach is offered in the AI for Gaming unit, preparing graduates for one of the most exciting and advanced AI applications today.
You can already see early stages of deductive reasoning by computers in a website tag cloud, where a software agent intelligently and clearly reports on the relevant importance of elements like text. It trawls everything from the content of a site to the searches that lead to it, updating the cloud with every new piece of data.
It’s the most recent incantation of the sort of data collection and ‘intelligent’ reporting deputy head of UWA’s School of Computer Science and Software Engineering Mark Reynolds looks forward to. “I want to introduce AI into the way computers do deductive reasoning so they’re a bit cleverer in arguing correctly and rigorously for the right answer” he says, “Checking every possible deduction systematically can be so time consuming we don’t get anywhere.”
One of the methods leading the AI charge is based on evolution, and Reynolds cites Lawrence J Fogel, the man who first applied Darwinian theory to complex problems. “Techniques like evolutionary algorithms — where solutions to difficult problems gradually improve when we let the proposed solutions compete, reproduce and mutate — are fun to work on and very clever,” he says.