A Journey to Learn About Artificial Intelligence
September 24, 2018
By: Rodney Lover
Avoiding an upcoming construction slow down, we veered off an early highway exit. As I relayed the GPS app’s instructions lighting up my iPhone, a coworker navigated our company van smoothly through the turns. Destination? A drop off location at a top reviewed café in Kitchener, just slightly out-of-the-way on a return trip from Toronto to London.
I bid my friend goodbye and spent some time in a nearby cafe. I had an hour or so prior to our SWO Angel spring learning session with our sister chapter in the area. Enjoying some fresh coffee, I caught up on email after a day out of the office, and took a few moments to look through our website analytics.
The day’s journey was simply full of Artificial Intelligence (AI) itself. On route, we used an incredibly advanced GPS app, listened to a tailored Spotify playlist, Googled and found the “best coffee roaster near me”. Upon arriving at the café, I reviewed website analytics composed of thousands stats. I am taking AI for granted daily and we have only scratched the surface of what is possible.
The SWO Angel event was excellent. We enjoyed food and drinks and networked with some likeminded investors who shared an affinity for startups. We were treated with a presentation by Professor Fakhri Karray, chair of University of Waterloo’s AI Institute and his colleagues, on you guessed it – Artificial Intelligence.
During Prof. Karray’s talk, a book was suggested “Prediction Machines” written by a group of Canadian economists. An interesting read, the book considers the origins of AI and speculates about possible advanced realities as the result of AI development.
Through all the complexities, what “Prediction Machines” helped convince me of, is that Artificial Intelligence is a convergence of inexpensive computing, new piles of digital data and the deep use of something we take for granted that make AI possible: “If” formula statements, which simply mean, “If this, then that”.
Certainly there are many levels of spreadsheet formula complexity, but think of AI as a robust and multifaceted Excel program with the ability of cheap, fast computers to process and act upon mounds of detailed data.
The power of this growing ability is that computers can learn by not only going forward in a pattern, but unlike humans, can be fed millions of points of previously recorded data and identify the patterns within.
Triage – AI improves front-line healthcare
Triage, a unique Ontario startup, came through SWO Angels earlier in 2018, pitching an app that can assess skin lesions and predict their level of danger. Imagine if you found something weird on your arm. Take a photo with your phone; upload it to Triage’s app, which then compares your lesion to all others in the database. The assessment is possible through thousands of images the Triage team was able to license. Each of which contained detailed recorded and linked diagnosis data that their AI processed.
Our issue as startup investors is that when companies come to pitch (especially tech startups), most currently relate that they have Artificial Intelligence in their offering. With due diligence and continual learning, it is incumbent on us to predict whether their level of AI is unique and something they can protect from competitors.
And we end up asking our own “if” statements... “if this (has AI), then will they succeed?”