NEXT Canada, GRIT Ed.06, Brain Gain vs. Brain Drain - Magazine - Page 9
mains such as health or education requires significant
additional training.
When we started Next AI in 2016, we were the only
AI-focused accelerator in North America. Thanks to AI’s
outstanding successes, this is no longer true: founders
have their pick of vibrant AI communities. In the previous edition of GRIT, NEXT CEO, Kyle Winters wrote
about aiming “to be an
employer of choice in
this competitive ecosystem.” Providing startups access to compute
would give NEXT and
other Canadian organizations a key advantage.
We have capable
and willing partners:
the CIFAR Pan-Canadian AI Strategy and
Canada’s three national
AI institutes have united thought leaders with
deep experience in compute for AI. Canada’s
New Digital Research
Infrastructure Organization is restructuring the
digital research landscape. But operating
our own infrastructure is
expensive and requires
specific expertise to
run and maintain. Meanwhile, most of our teams
have access to some “free tier” credits from US-based
cloud providers and spend out-of-pocket when those
are exhausted.
Gibson and Bodkin stress that it’s “important to not
default to a commercial cloud solution.” I agree, the optics of subsidizing out-of-country cloud compute are
very poor and do nothing to grow needed Canadian
infrastructure. Furthermore, startups may be working
with data that for reasons of privacy or security needs
to stay in the country. Turning to the burgeoning Canadian commercial cloud sector could be the best option
for startups and accelerators.
At NEXT, we often use “dogfooding” within the accelerator, where startups become the early adopters of
products and services of currently or recently graduated startups. By facilitating relationships and pilots between Canadian cloud providers and AI startups with
growing compute needs,
we support a national infrastructure that benefits
researchers and government organizations who
increasingly leverage AI to
drive economic and social
benefit for Canadians.
A new generation of
privacy-enhancing
technologies (PETs) has the potential to support collaborations among Canadian
SMEs where previously not
possible due to regulation
or competing interests.
For example, these technologies can leverage the
effects of larger datasets
without needing to pool or
share data outside organizations. Vector’s health-AI
researchers have used PETs
to enable multi-hospital
collaborations. A stronger,
sovereign cloud is the backbone for such partnerships
in the commercial space.
The Canadian research community through the
Pan-Canadian AI strategy is expanding Canada’s critical capacity in compute. By working together we support our AI-enabled startups who desperately need this
resource. Partnering commercial and noncommercial
infrastructure will ensure that Canada is the clear winner for founders deciding where to start or scale.
By facilitating relationships and pilots
between Canadian cloud providers and
AI startups with growing compute needs,
we support a national infrastructure
that benefits researchers and government
organizations who increasingly leverage
AI to drive economic and social
benefit for Canadians.
Insights for navigating a
rapidly changing world
Disruptors, an RBC podcast, is an ongoing podcast series hosted by
SVP John Stackhouse about reimagining Canada’s economy in a time
of unprecedented change. It features thought-provoking conversations
with Canadian business and innovation leaders about planting the
seeds of a new economy.
rbc.com/disruptors
Listen wherever you get your podcasts
Acknowledgement: This piece was prompted by conversations with Ron Bodkin (VP of AI Enginnering and CIO,
Vector Institute), Tomi Poutanen (Head, Layer 6 AI; Co-founder and Advisor, Vector Institute), Cameron Schuler
(Chief Commercialization Officer and VP Industry Innovation, Vector Institute), and Michiel Vos (Manager of AI
Startups and Scaleups, Vector Institute).
8
GRIT