By Najah Hassan & Dale Gottlieb, Cannon Senior Editor & Cannon Editor-in-Chief

Machine Learning or Machine Dumbing? Credit Muhammad Ali

The “I’ll-Figure-It-Out” Computers: Najah Hassan

Starting in September 2018, the Division of Engineering Science will introduce a new option in Machine Intelligence. While some students are excited for this new addition to the curriculum, others worry that this proposed program may be too far apart from the definition of ‘engineering’. Until about a century ago, an engineer was more commonly associated with the study of mechanical structures, civil infrastructure and material composites. However, today engineering also includes electrical engineers, software engineers, financial engineers and now, machine intelligence engineers.

 

Machine intelligence or machine learning is defined as the study of developing machines that can think for themselves. New developments in the field have shown that this is a growing area of interest with a huge demand for graduates with this specialty. The new option through the Division of Engineering Sciences proposes to equip students with the ability to use multi-disciplinary wholesome thinking to solve complex problems in the world. Why is this so important? The amount of data that is being collected through our devices on a daily basis is growing and humans just are not able to keep up with the organization and analysis of this information. By teaching computers to find patterns in data on their own, we can solve some complex problems and find the answers to questions that we had not thought of previously.

 

Take the example of 23 and Me’s new research towards discovering the cause of Parkinson’s disease. 23 and Me, a personal genomics company, is using data mining techniques to sift through around 2 million genetic samples that it received from its consumers. The company has used this information to find more than a dozen different mutations that are linked with the disease. Doing further analysis in this realm could lead to the cure for Parkinson’s, allowing us to break through to new domains of healthcare.

 

The applications for machine learning are numerous. Toronto’s recently launched Vector Institute promises to use artificial intelligence to improve the lives of Canadians and establish economic growth by focusing on the principles and potential of machine learning. It aims to be Toronto’s new research hub and has a lot of opportunities for graduates in machine learning.

Self-driving cars, machines that find cures, and future predictions based on past performances have all become a reality. The world is moving towards developing technologies that lead to faster, smarter and more efficient societies. As engineers who aspire to make a difference in the world, a knowledge of machine learning and artificial intelligence in this day and age will be a useful skill to have to adapt to this new change. Engineers can combine their problem-solving skills, mathematical background and engineering design strategies with machine learning to solve some of the world’s biggest problems. No doubt, there is a lot of work to be done in this field!

 

The “I’ll-stop-you-from-figuring-it-out” computers

By: Dale Gottlieb

From the Stone Age to the Iron Age, the Bronze Age, and the Silicon Age, the technological development of society has always been measured by our capabilities to manipulate the materials around us. We need a fundamental understanding of nature if we ever wish to achieve the great strides in the quality of life that engineering rightfully takes credit in developing. Even the computer and the software that it runs only exists because of the knowledge to manipulate silicon on the atomic level and to tailor it to our needs. This is why I think it’s a shame that engineering is drifting away from the physical sciences and towards software.

The new introduction of Machine Learning engineering in Engineering Science at the loss of infrastructure engineering I feel is a grave representation of the stagnation of progress. Engineering Science used to be defined by the broad choices in studies available. For old-school students, there was Infrastructure Engineering, for students interested in chemistry, there was Nanotechnology Engineering, and so on. Now there’s Aerospace Engineering, Robotics Engineering, Electrical Engineering, and Machine Learning…Engineering (?) all teaching a similar subject matter.

Other disciplines are following suite. Materials Engineering, once called Metallurgical Engineering, changed its name to cover catchier fields of research like nanotechnology. Mechanical Engineering now focuses on circuit design for robotics, and Industrial Engineering focuses on computer optimizations.

The uniqueness of engineering is being lost, and the ability for engineers to do what scientists can’t is fading. A condensed matter physicist is better at nanotechnology than a materials engineer, much like a Computer Scientist is better than a Machine Learning Engineer at programming. Soon, the only group of people with the knowledge and ability to progress the basic knowledge of science will be lost to the promise of high paying, immediate payoff jobs in Silicon Valley.

At the present, especially as young millennials, it feels as though an app like Uber is revolutionary, but in reality, it adds nothing to society. The ability to hail a taxi a second faster using an app pales in its affect on society compared to the design of the car, or even a small but essential component of a car like the transmission. A field of engineering like Machine Learning Engineering will get many students high paying jobs in Silicon Valley, but will detract from the development of society engineering has held so dear since its inception.

To some, my argument against the introduction of Machine Learning Engineering might be translated to ‘machine good, machine learning bad’, but I think it stems to a much bigger issue than we can predict today. With the increasing reliance on computers to solve our problems, and the decreasing number of people researching anything else, we’re destined to be stuck in the silicon age forever. As the top engineering school in the country, I feel UofT needs to think twice before setting an image for all other schools that the only thing an engineer is good for, is what a computer scientist is great for.

Leave a Reply

Your email address will not be published. Required fields are marked *