Could an innovation like ChatGPT see the light of day in Canada? It seems unlikely. Not only are Canadian AI companies struggling to retain their researchers, but the vast majority of intellectual property created here is owned by American and Chinese companies.
In short, a Canadian ChatGPT is unlikely to see the light of day. Canada still has a leader in the development of AI similar to ChatGPT. Toronto-based Cohere is an expert in large language models, or “LLMs,” as they’re called in English. Its problem: it is underfunded compared to its counterparts in other countries. Its main backer is Salesforce Ventures of California. Cohere raised $165m earlier this year. This is very little compared to the 1.3 billion US $ of Anthropic, an AI laboratory created by former OpenAI, which is on its side with 11 billion US $ in venture capital raised in recent months.
The widespread lack of interest in AI and new technologies among Canadian business leaders is largely responsible for this imbalance, observes Scale AI. The organization that oversees the Canadian AI supercluster demonstrates this in a report published at the end of March, entitled Portrait of AI from here.
“There is a severe lack of domestic demand for artificial intelligence applications,” notes Clément Bourgogne, chief operating officer of Scale AI. “We have some of the best researchers in the world, but there is no room in the private sector to create a growth space for their technologies. »
For businesses, the most direct way to innovate is to invest in R&D. And while business spending on R&D has been on the rise in nearly all of the world’s developed economies since spending by Canadian businesses is down. Scale AI estimates that the total value relative to Canadian GDP of these expenditures in the country fell by 2.8% between 2015 and 2019. It increased between 2.2 and 8.3% in the other countries.
The real problem is that few of our researchers work for the benefit of a Canadian company. The gains currently achieved do not serve the Canadian economy.
Canada’s private sector is too dependent on government support for R&D and new technologies, Scale AI believes. “It’s once companies see the potential to increase their sales or reduce their costs that they invest. Before, when taking the first risk, they hesitate a lot, ”explains Clément Bourgogne. Scale AI has just raised $125 million from Ottawa to precisely stimulate the faster adoption in Canada of AI applications produced in the country.
China Branch
When it comes to AI, Canada is a huge producer of intellectual property. It is the fourth country where the most results of investigations are published. Canadian AI researchers are reaping more than their fair share of patents for their inventions.
However, half of these patents are won by foreign companies. IBM and Microsoft are by far the two largest holders of intellectual property in Canada. The next three are Chinese organizations: the State Grid Corporation of China, a state-owned company, the Chinese Academy of Sciences and Xidian University in Xi’an, a city located in the heart of China.
This is a worrying situation. The West increasingly fears China. In Ottawa, she is suspected of interfering in the operation of the electoral system. In the strict economic sense of the question, the strong representation of both Chinese and foreign companies more generally should also sound an alarm, says Clément Bourgogne. He insists on the importance in the coming years of increasing the presence of AI researchers within Canadian-owned companies.
“I don’t know if China is more of a danger than any other country. The real problem is that few of our researchers work for the benefit of a Canadian company. The gains currently achieved do not serve the Canadian economy. »
Brain drain
Driving more Canadian companies to hire AI professionals is not so simple. At least Canada attracts students in huge numbers. The Canadian strategy is simple: if they study here, then they will want to work here.
However, this is not necessarily the case. Canada certainly attracts students and young researchers, but the professionals they subsequently become tend to leave the country for the benefit of the United States or other countries. Salaries in Canada are up to 33% lower than those offered in the United States or Israel, notes Scale AI. The challenges that researchers are called upon to meet are also more alluring there.
“Specialists are attracted to companies where they will be able to solve complex problems,” Scale AI writes in its report. “If the Canadian market doesn’t provide such opportunities, it’s highly likely that they will move to countries where the AI program is pushed further. »
Scale AI still sees a glimmer on the horizon: local companies as disparate as McCain, Kruger and Pratt Whitney have begun to value AI in their growth. But more is needed, concludes Clément Bourgogne. “We need to better demonstrate that the large companies here are also creating value and have ambitious projects for our AI researchers. »