Vector Institute

What it is

The Vector Institute for Artificial Intelligence is an independent, non-profit AI research institute in Toronto, Ontario. Its focus is deep learning and machine learning.^[1] Under the Pan-Canadian AI Strategy it is one of three national AI institutes, alongside Mila in Montreal and Amii in Alberta.^[2]

Origins

The institute launched on 30 March 2017 at the MaRS Discovery District in downtown Toronto.^[1] It launched in partnership with the University of Toronto, and its founding scientists were affiliated with the university.^[1] Wikipedia lists Brendan Frey, Geoffrey Hinton and Raquel Urtasun as founders.^[1][2]

At launch its funding came to $50 million from the Province of Ontario, more than $80 million over ten years from over 30 companies, and a share of the federal government's $125 million Pan-Canadian AI Strategy, which is run with CIFAR.^[1] This mix of public money, corporate money and a national research programme is what set the institute's early shape.

Geoffrey Hinton was a founding scientist and served as its Chief Scientific Advisor.^[1][2] As of June 2024 Hinton was still Chief Scientific Advisor, and Daniel Roy was Research Director.^[2] The University of Toronto describes Hinton as Vector's co-founder and chief scientific adviser.^[3]

Known for

The institute's stated focus is deep learning and machine learning. As of 2023 it counted 143 faculty members and affiliates, 38 of them CIFAR AI Chairs, along with 57 postdoctoral fellows and 502 students.^[2]

Leadership has changed hands once. Garth Gibson became its first president and CEO on 2 January 2018, and Tony Gaffney succeeded him in 2023.^[2]

Place in AI history

Vector's significance lies in its national role and its partnership with a major university. It sits in the Pan-Canadian AI Strategy alongside Mila and Amii, and it brings together a provincial government, many companies and a research university in one organisation.^[2][1]

In 2024 the institute moved from the MaRS Discovery District to the Schwartz Reisman Innovation Campus.^[2]

For a reader, the institute is a useful example of how national AI programmes were organised: a research institute, a set of public and private funders, and a leadership team whose members stayed tied to a university.