Looking to get artificial intelligence off the ground in your organization and become your company’s first AI exec? Not so fast. These experts say there are a couple of questions you should be asking yourself to make sure your pet project doesn’t get out of hand.
Does artificial intelligence (really) have value for your business?
“AI is the new ‘innovation lab,’” says Automat.ai CEO Andy Mauro, poking fun at the current hype. Element AI Senior Manager of Industry Solutions, Richard Zuroff, says machine learning should only be applied if:
- A task or set of tasks rely on a lot of intuition (recognizing a face, for example)
- Making the best decision requires different types and significant amounts of data
- It simplifies decision-making
- It makes interacting with a system more organic
- It improves a core process, generating a competitive advantage
“As an executive, the last thing I want to get into is a 100% science project.” — Andy Mauro
What’s your problem?
If your company meets at least one of the above conditions, congratulations: AI can help. Now, it’s important to precisely define the problem you want to solve.
A simple problem
You’re a retailer trying to plan seasonal hires.
An AI solution
Estimate online transaction volume from November 1 to December 20 based on the five previous holiday periods to determine how many temporary warehouse workers to hire.
Most companies will want to consider AI on a macro level, solve multiple problems or automate several sets of tasks. However, Stradigi AI Co-Founder and Chief Scientific Officer, Carolina Bessega, says it’s wise to start small and build incrementally.
Buy, build or partner?
Another question arises: Should you purchase a solution, bring AI research and development in-house, or partner up? There’s a case to be made for each scenario.
Listen to the McKinsey Podcast report on Artificial intelligence in business: separating the real from the hype.
Buy: “If the solution you need exists, buy. If not… go on that journey,” says Shelby Austin, National Service Line Leader, Strategic Analytics and Modelling at Deloitte Canada. As long as risk, cost, experience and revenue are a go, you should build it.
Build: Building your own capacity comes with a caveat: you must hire more than one AI researcher. “An important part of it is collaboration. It’s difficult to make strides when you’re the only researcher in a room,” says Carolina.
Partner: Carolina also suggests partnering with an AI company that has research scientists and engineers — it’ll get you results faster.
Read more from the “Back to business” special:
The C2 Montréal Minutes: Actionable insights for creative business leaders
This article is excerpted from the upcoming Transformative Collisions: The C2 Montréal 2018 Minutes, available for your reading pleasure this fall at c2m.tl/minutes2018.
Subscribe to our newsletter now and we’ll let you know as soon as The Minutes are out.
Questions or comments? Drop us a line at email@example.com