“Communitech’s Tech Leadership Conference brings the best entrepreneurs, innovators and thought leaders [together] to connect, learn and help shift the power and control of technology in the right direction.”
That’s a pretty big mouthful. Said another way, the goal of the conference is to bring together a bunch of smart people who are interested in technology and hope cool things happen.
There were three reasons for a representative of Capacity Canada to attend the Tech Leadership Conference. The first reason was to be seen. Many conference attendees are strong, influential leaders; they are helping to define the direction of this region and, in all seriousness, the world. Capacity Canada is a part of that movement and we need to assert that role with confidence.
Second, the topics of discussion at this year’s conference should be of interest to Capacity Canada. As an organization, we always need to consider scaling our existing programs, but also how we can build growth into technology platforms that we may create in the near future. The role of data is becoming increasingly important in the social sector and its ethical gathering and use needs to be built into projects from the outset.
Third, we need to continue to build relationships with local tech organizations, especially in the for-profit sector. As the sort of projects Capacity Canada undertakes changes and diversifies, how we approach partnerships (for both funding and implementation support) is critical to their success.
When originally announced, this year’s conference was supposed to have changed to include an additional day for longer form workshops. Workshops were scheduled to learn innovation techniques and design sprint methods (led by Mark Connolly, who facilitated a Capacity Canada design sprint in February). However, there was not enough interest in the workshops and they were cancelled.
The conference took place on Thursday, May 11 in Kitchener, Ontario.
This document is intended to capture the spirit of the conference and how its discussion applies to Capacity Canada.
As in past years, the event was held in Kitchener at Bingeman’s Conference Centre. While this is hardly an inspiring venue, it is large and reconfigurable, easy to get to, and parking is plentiful.
During registration, attendees were provided with nametags and a buffet breakfast. During breaks and lunch, food and drinks were provided.
Attendance seemed healthy, with sessions often filling up to capacity. Professional or business casual attire was the norm and reflected the cross-section of attendees (tech management, professional services, etc.).
There were five areas used by the conference. A large conference hall was used for keynotes. Three smaller halls were used for topical speakers. Breaks, lunch, and what Communitech called the ‘Cool S#!t Zone’ shared a large common area. Demonstration booths were set up to show off local products.
Breaks were on the longer side, but this was useful for networking and catching up with colleagues.
Communitech provided a custom app for your phone in order to register speaker questions, notify attendees of schedule changes, etc. This worked out fairly well and use seemed high.
The MC for the day was Sarah Prevette, a serial entrepreneur and sought after speaker in her own right.
The event was opened and closed with comments by Communitech CEO Iain Klugman. This year is the 20th birthday for the organization and Klugman reflected upon the changes that have occurred since its inception (by 43 founding companies).
Tom Jenkins, co-founder of OpenText, also spoke briefly about his book, Ingenious (co-authored with Governor-General David Johnston). This book, together with another entitled We Built This: Volume 1, were given to all attendees.
— Ali Asaria
Ali Asaria was the first speaker of the day. He told the story of going through school for computer science, becoming disenfranchised with the corporate world, and branching out to do his own thing (first at Well.ca, and now with Tulip Retail).
His talk focused on reframing the discussion around the loss of jobs as a result of disruptive technology. Artificial intelligence (AI), robotics, and automation are all poised to dramatically change how humans work. Will, as some predict, entire classes of job be eliminated overnight? Will we soon have, a ‘useless’ class of citizens? Or, as he proposed, do we need to look at the potential shift as an opportunity?
Asaria’s work at Well.ca left him conflicted. Within the span of a few short years, Well.ca became the largest online pharmacy/healthcare store in Canada and, as a result and without question, competitive retail jobs were lost. What did those people do? What will happen to the 8M+ retail jobs in North America when AI, robotics, and automation reach a tipping point?
Asaria gave an example of how we’re currently in the middle of the mess. We crave automation yet we value human craft: we’ll happily use a pod-based coffee maker…but we also enjoy going to a coffee shop to get a barista-crafted espresso.
His proposal? We, as a society, need to look at which tasks can be performed by the ‘three horsepeople of the apocolypse’ (AI, robotics, automation) and how can we fill that void with better human experiences. We need to invest in people and in their craft.
It is hard to argue against investments in people and their skills. Time will tell how dramatic the impact of AI, robotics, and automation is to broad classes of workers. As Asaria said, “jobs aren’t going away, but the work is changing”. How quickly that transition occurs also plays a large role in whether workers will be displaced or if younger generations merely train for different types of work.
On the topic of basic (or guaranteed) income, Asaria had an interesting (if overly simplistic) point: people derive considerable value from their jobs. If, as a result of disruptive technology, classes of workers are out of work, providing them a basic income will do nothing to replace that value; we shouldn’t just give them money and label them useless.
— Patty McCord
Patty McCord delivered the opening keynote. McCord is well known from her tenure at Netflix as the Chief Talent Officer. Her talk was entitled ‘Power, not empowerment’ and collected anecdotes, lessons, and criticism from the tech corporate world.
She described her philosophy that companies aren’t families, nor should they be. Instead, they are teams and leaders are coaches — motivating people to greatness within the corporate world is no different than motivating hockey players on the world stage.
Amidst the funny stories she told (which were nothing more than that), there were a few ideas to learn from and take away.
First, when thinking about your organization’s culture, think about whether you want to be a great company to be from. In other words, would someone look at a CV, see your organization’s name, and be impressed or want to know more? This simple test encapsulates lots of positive attributes, but also acknowledges that every employee will eventually move on, and there are lots of reasons why that is completely healthy.
One of the reasons someone might move on with the support of the organization leads to the second interesting idea proposed by McCord. Employees should ask themselves: “Is what I love to do, what I’m really good at doing, something [the organization] needs someone to be good at?” Organizations change, situations change, people change…admitting that like full formed adults and moving on is in everyone’s best interest.
The third take away was a mental exercise McCord recommended in order to not only visualize and plan for the short to medium term, but to more objectively critique your current situation. Imagine your organization in six months: What exactly does tremendous success look like? Who is there? What are you talking about? What are you struggling with? Are there more meetings? Fewer meetings? How has your day changed?
Next, compare that six month picture of great success to today: What is really different? By working backwards and contrasting great success to today you can visualize what needs to change now. This technique can be used not only for hiring people, but for funding pitches too — sell the six month vision, confident in the knowledge that you now know what needs to change.
The final point from McCord cannot be emphasized enough: tech people follow problems. Great tech people got into the industry because they love solving problems; give them one.
— Tonya Custis and Benjamin Alarie
The use of artificial intelligence (AI) in the legal field seems a bit academic. However, the talk by Tonya Custis and Benjamin Alarie was engaging and provided a glimpse into the practical applications of such technology.
Custis is an academic researcher working for Thompson Reuters and develops ways to use AI to help the legal sector. In particular, she studies linguistic analysis and the breaking down and interpretation of text.
She stated that AI has been around for decades but the recent explosion of fast commodity hardware, large digital data sets, and commonly available tools make it useful in more arenas. While legal tax documents provide a great sandbox within which to play today, other industries will be able to leverage the technology as time goes on.
Alarie is a lawyer and CEO of a startup whose goal is to predict the outcome of tax cases using AI. By ‘teaching’ the software to understand the inputs (legal cases) and outputs (outcomes of those cases), the AI can be used by a lawyer to decide if it makes sense to go to trial.
Both speakers were excited about the future of AI as a means to augment decision making in a well structured, well documented industry. When asked how such technology might translate to other industries, Custis had a few rules of thumb to determine feasibility.
First, can a human do the work in a short period of time? (e.g. less than one minute) Of course, reading through tax documents and deciding if it makes sense to go to trial may take a human more than a minute, but each constituent step may not. By breaking the larger task into smaller tasks and examining them individually, so can the feasibility of the larger task be considered with respect to AI.
Second, is there a large enough data set to be able to train the software? In the case of law documents, the data set is well defined, digitized, and — wonderfully — maps inputs to outputs/outcomes.
Alarie indicated that their AI software was getting more than 90% success rates predicting the outcome of cases. This was tested by identifying a specific tax law scenario, gathering 1000 relevant cases from literature, training the software on 700 of these cases, then testing the trained software on the remaining 300 cases.
I couldn’t help but think that such a system would be wonderful to predict whether, given a set of inputs/variables, a given funding proposal would be accepted or not. Such a system could not only help funders better describe the acceptance criteria, but reduce inefficiences on funding teams within social organizations.
The speakers and session were introduced by Joseph Bou-Younes from the Open Data Exchange (ODX). On the same day as the conference, the Waterloo ODX office (located in the old police station in Uptown) was opened, with anchor tenant CIBC. ODX is funded by the University of Waterloo, D2L, Canadian Digital Media Network (CDMN), OpenText, and Communitech with matching funds from the Goverment of Canada.
— Cheryl Ainoa
Cheryl Ainoa is the COO at Kitchener-based D2L. A common theme throughout her career has been scaling: she’s started companies and grown them to acquisition and she’s worked with companies like Yahoo! as they grew from 3000 to 18000 (in three years!).
Her talk described some of her techniques for dodging scaling issues, as well as how to run a business with an eye for growth. She identified three common elements that occur with successful scaling: change and the ability to handle it properly, discipline when developing a product, and being able to share collective knowledge within an organization efficiently.
As an organization starts out creating a product or service, a larger percentage of resources can be applied to research and innovation. However, as customers come into play, support and maintenance of the product and customer reduces that ratio. The key is to be able to say ‘no’ to customers. Rather, identify your priorities for the product/service and organization and stick to your plan.
When developing that product over time, discipline is required not only to say ‘no’ where appropriate, but also to frequently assess its place relative to the market. She described (albeit not by name) the Kano model, used within product development circles to identify the core requirements for a product, the features that excite customers, and which missing features would prevent a sale.
Sharing knowledge within an organization is obviously something Ainoa (and thus, by extension, D2L) takes very seriously. Not only does sharing information reduce time to develop features, it establishes good patterns within the organization.
At D2L, many employees are expected to attend professional conferences that have been shown to be practical (“Is this just a party for developers?”). In order to make the most of these conferences, employees are given tasks to complete (e.g. talk to three people about technology XYZ, add them to your LinkedIn network, and ask them a question once a quarter).
Taking a look at how other organizations are doing similar tasks and tackling similar technical projects is always a good idea. Remember to place a competitive organization’s decisions into the proper context (they may have made that change for reasons that don’t apply to you). One practice that I’ve seen work really well that Ainoa championed was field trips — take a team out to watch how a customer or other organization does something!
— Billy Beane
Closing out the conference was Billy Beane, EVP of Baseball Operations with the Oakland A’s. He is best known as being the subject of the movie Moneyball (he was played by Brad Pitt). Together with his assistant, Harvard economist Paul DePodesta, Beane revolutionalized how small market baseball teams evaluated performance in players through the use of commonly available statistics.
The Moneyball method used not only statistics but recognized that the value a player provided a baseball team was the ability to get on base, independent of how they looked as a ball player (“the eye test” — or all the other intangibles traditional scouting considered). Using Moneyball, the Oakland A’s were able to find incredible value in players who were being overlooked by other teams.
Beane’s talk was engaging and entertaining. He spoke about how the book and then movie came to be. He spoke about the battle involved with challenging the status quo (and, with baseball, tradition — even with scouting — is sacrosanct). He spoke about how, within 20 years, scouting and the idea of value has changed completely, all as a result of data and its analysis.
From a data science point of view, several points were of interest. First, Beane acknowledged that the system they used wasn’t new. Rather, they looked at existing systems from a different point of view and made them work for their situation. Data can be used in different ways, present different biases, and can tell a different story…depending upon how it is approached.
Second, all data isn’t the same, even when it is. Beane explained how statistics related to hitting a ball into left field under a given circumstance was shown to be very effective at getting on base. However, if a given hitter was caught out in that same situation, the statistics still show a fly out. In other words, that was the right play given the situation, but the statistics record it as a failure. Oakland was able to look into the context for a given statistic and mine additional information to their benefit.
Third, by setting aside biases and history for a given process, there were measurable gains. Only 3–5% of high school and college players signed by a team (yes, they get signed very early) will make the major leagues. However, since high school players have significantly less data around them (vs. college players), they are more of an unknown. The A’s stopped signing high school players, instead relying on data for college players exclusively. The result? A 13% success rate.
Perhaps the most human aspect of Beane’s story was the change in culture within professional baseball. Whereas baseball used to be run by ex-players or family members of owners, etc. there is a shift towards hiring people who love the sport and understand the science involved. He told stories of receiving resumes from astrophysicists out of Cornell and Oxford, Cal. Berkeley Economics doctors, and so on. Baseball has become a meritocracy.
So, while the Oakland A’s are still a small team in a small market, the use of data, the application of critical analysis, and keeping an open mind have allowed them to find value where others haven’t. They’ve changed not only their own culture but that of an entire sport.
The 2017 Tech Leadership Conference seemed to be suffering from an identity crisis. The topics being discussed were certainly timely — data, robotics, cybersecurity, artificial intelligence are all ‘hot’ right now — but the content associated with those topics were too high level to be universally practical. Was the goal of the conference to inspire? To empower? To tool up? To entertain? It was unclear.
But how did this conference relate to Capacity Canada? While aspects of the topics were applicable in a general way, the projects currently underway for Capacity Canada don’t need AI or robotics or likely even big data (yet).
What was useful, however, was hearing about how others have wrestled with the human element (Asaria and disruptive technology rendering classes of workers ‘useless’), struggled with focused scaling (Ainoa and the need to say ‘no’ to projects and features), and how culture will always define an organization (McCord and her war stories).
An an attendee, the conference was well run and provided opportunities to meet new people, explain to them why I’m excited to work with Capacity Canada, and get feedback on some of our ideas. In that respect, attendance at the Tech Leadership Conference was a success.
n.b. At the end of the conference, Iain Klugman said that while this was the tenth year the Tech Leadership Conference had run, it would be its last. Instead, in the spring of 2018, Communitech will be hosting Communitech 151, a ‘large-scale, urban conference’. Details were scarce, so time will tell if it is in Capacity Canada’s best interest to attend.
Tech Leadership Conference
Canada’s Open Data Exchange (CODX)
Communitech conference wrap-up article
Communitech 151 (TLC conference replacement)