Two of the world’s most influential artificial intelligence leaders took the global stage this week, pushing for a collective effort to delay the rollout of a technology they believe could spiral out of control and imperil humankind.
uncertain
they → take → humankind
Not so fast, says Canada’s AI Minister Evan Solomon.
asserted
Minister → say → ?
“We are absolutely clear that we are in support of safe models … and that models should not be recklessly released,” Solomon told the Star in an interview.
asserted
Solomon → release → interview
“But we also have to be careful that big tech companies are not setting up a system where they are governing themselves with their own guardrails.
asserted
they → have → guardrails
Our country is going to set the rules to protect our citizens, and the guardrails and governance systems are going to be set by democratic governments whose accountability is to citizens, not big tech companies, who are both building the systems and now suggesting ways to create their own guardrails.
asserted
who → go → guardrails
Solomon’s unwillingness to immediately back leaders like Anthropic’s Dario Amodei and OpenAI’s Sam Altman comes as nations around the world grapple with finding common ground on what those guardrails should look like.
Prime Minister Mark Carney was one of 22 world leaders to sign on to a United Nations declaration this week calling for AI companies to develop transparent safety and testing protocols for advanced models, for governments to establish unified standards, and for UN members to back an international body to oversee a monitoring regime.
Several of Canada’s G7 allies opted not to sign, including France, Japan and the United Kingdom, while the United States and China — the superpowers competing to dominate on frontier AI — also left their names off the pledge.
asserted
States → back → pledge
In recent days, both Canada and France have convened meetings with G7 members and other countries to discuss accelerating global co-operation on overseeing AI, including Carney’s pitch for a “technology stability board.”
asserted
Canada → convene → board
“It’s a model based on the Financial Stability Board that regulates financial markets.
asserted
that → ’ → markets
That would be the best way to also oversee global AI deployment,” said Solomon, who added that Carney and French President Emmanuel Macron are working closely on the proposal, which has no release date planned.
asserted
which → oversee → date
“There are some key areas that we are looking at.
asserted
we → be → that
Model evaluation and testing, model deployment, incident reporting, co-operation with our safety institutes,” the minister said.
asserted
minister → say → institutes
Solomon said his American counterpart, U.S. President Donald Trump’s science adviser Michael Kratsios, was unable to attend either meeting, and that Ottawa has had “no outreach” with China on holding AI companies to account.
asserted
Ottawa → say → account
But he said he is “watching extremely closely” this week’s meetings between Trump and Chinese President Xi Jinping, where the U.S. president said “superintelligence” would be among his top priorities.
asserted
superintelligence → say → priorities
“The decisions we make in these areas today can promote peace and prosperity for decades and decades to come,” Trump said Thursday.
asserted
Trump → make → decades
OpenAI’s Altman, who was expected to attend Trump’s state dinner for Xi Thursday evening, recently said the pair of leaders could win the Nobel Peace Prize if they are able to reach a deal on AI.
uncertain
they → expect → AI
Altman and Amodei both addressed the UN Security Council Wednesday, where they called for unity on monitoring the potential harms posed by advanced AI systems.
asserted
they → address → systems
“I see two main categories of risk.
asserted
I → see → risk
The first is misuse of AI models, for example, misuse by bioterrorists to create biological weapons.
asserted
first → create → weapons
The second risk is loss of control, model capabilities accelerating beyond developers’ ability to control them,” Amodei said.
asserted
Amodei → accelerate → them
“If managed poorly, I even believe that AI could be a risk to humanity as a whole.”
uncertain
AI → manage → whole
Altman said it doesn’t matter where people put the risk of catastrophe.
asserted
people → say → catastrophe
“None of these levels are remotely acceptable, and we should not train models (for which) we cannot make an extremely strong case they will be able to keep under human control,” he said.
asserted
he → train → control
The leaders spoke after Yoshua Bengio, a Canadian AI pioneer, addressed the assembled members.
asserted
Bengio → speak → members
“Citizens and governments should not accept the companies’ claim that they are trapped in a race.
uncertain
they → accept → race
It is the product of choices, choices made by the companies themselves.
asserted
It → make → companies
It is also a race where everyone loses,” Bengio warned.
asserted
Bengio → lose → ?
Another of Canada’s AI leaders, Cohere CEO Aidan Gomez, recently questioned Altman and Amodei’s motives for raising the alarm about frontier AI and urging for slowdowns.
asserted
Another → question → slowdowns
“Here is the question nobody is asking clearly enough: should a handful of select, market-dominant AI companies from Silicon Valley get to define the rules and safety standards of a generational technology for the entire world?
asserted
handful → ask → world
All while simultaneously determining how fast this technology progresses?” Gomez wrote in a company blog post earlier this month.
asserted
Gomez → determine → post
“Once again using fear under the pretext of protecting the public, these oligopolies are now requesting to bend competition rules and be permitted to dictate the terms for everyone else.
asserted
oligopolies → use → everyone
Solomon said he won’t be waiting for tech titans to build models and dictate the terms of their governance and use.
asserted
titans → say → governance
“That’s not how it works here,” he said.
asserted
he → ’ → ?
“This is the purview of sovereign governments, and that’s why we’re acting as we are.”
asserted
we → ’ → governments
The Aporia
