CBC’s David Common talks to Francis Syms about concerns over the development of AI

CBC English
16 followers
3 Video Views·Sep 15, 2026
AI Generated·May contain errors
Video Summary
AI GeneratedExperts and policymakers are warning that rapid AI development could pose an existential threat to humanity. The discussion centers on the need for mandatory \"kill switches,\" the dangers of a profit-driven race for superintelligence, and the urgent need for democratic regulation to prevent a small group of Silicon Valley companies or foreign adversaries from controlling the technology.
We've heard repeated warnings this month that AI development is happening so fast that there
is a risk it could wipe out humanity within the decade.
One of the co-founders of the AI giant, Anthropic, is calling for mandatory kill switches.
Jack Clark explained the potential danger to the BBC.
The basic risk is, how well do you fancy your chances of being able to deal with the risks
of a system that is 100 times smarter than the smartest person on the planet?
I think that your chance of dealing with that and getting out in a safe way is pretty low.
So, high level, you need to control AI as a technology because you're making things that
might be smarter than people. Things that are smarter than people could be really,
really, really dangerous.
The backlash is actually bringing together two unlikely bedfellows, U.S. Senator Bernie Sanders
and conservative activist and former Trump operative Steve Bannon.
They are set to share the stage at an event in Washington today,
calling for Congress to work faster to curb the impact of AI.
To help us understand, Francis Sims, Associate Dean of Information and Communications Technology
at Humber Polytechnic joins us now. Francis, good morning to you.
Good morning, David.
The assertion here is that AI is happening, the development is happening so fast that it will
at some point get uncontrollable. That it may go rogue and therefore could threaten
humanity in some fashion. How is it that the companies that are developing it themselves
can't build something in now to deal with that?
Yeah, so there's this culture within these companies, they even refer to it as model
fatigue, where there's this hypothetical race against likely China and other countries around
the world to build the fastest and most capable model first. Because those that build it first
will be the ones that bring the greatest returns to the investors. We have to remember that these
are market companies that are focused on profits for their investors. So they might have altruistic
goals, but number one, they want to win the AI race. And so part of that means that they have to
release model after model very quickly. And unfortunately, what happens sometimes
is we sacrifice things like quality control and testing.
The challenge, as you say, there is a race against China, which has not quite the frontier
models, but very close to the frontier models at a fraction of the price. Trump rejecting calls
for any kind of regulation, he calls them a hoax. And he is claiming that the only
thing that should be touted is whoever wins AI wins in the race over China.
I wonder where your sense is on whether there is even the possibility of regulation?
Yeah, I 100% think there is. I think the race is really towards trustworthy and ethically
responsible AI. And I'm here in Vienna at the UN Atomic Energy Conference this week. And that's
where the focus is on transformative technology and how it can benefit people and we protect
against the harms. So the world has done it before. But what we needed to do is take that technology
out of the companies and make sure that it was in the hands of the democracies and the countries
around the world. And the problem here, and Aidan Gomez, who's the founder of Cohere out of Toronto,
said that the behavior of the Silicon Valley, and you can read this on his blog on his website,
is akin to cartel behavior. Because the problem is, it's a small group of companies out of Silicon
Valley that are dictating the rules of this game. And that's not good for the rest of the world.
The average age in the US Senate is I think 66 right now. And when you look at hearings,
the very few hearings they've had about AI, it hasn't always seemed like senators even
understand the technology. What is the prospect that they could regulate at all?
Yeah, well, I think there's now two bills or acts that are passing,
or being introduced or starting to get discussed. One is the AI Kill Switch Act. And the focus of
that is making sure that there's an off button, right? So that if AI does get to the point where
it's out of control, you can pull the plug. And then the second one that's being introduced by
Sanders, and Sanders is not a young fella himself, is focused on banning super intelligence and
pausing development of advanced AI until regulations can be put in place. And so I'm
encouraged, you know, Bannon and Sanders, and there's many young people as well. But if they're
able to get their minds wrapped around it, right, and they're a little bit older, then they can
communicate and translate that for their peers. And so I'm very hopeful to see and they're on
opposite sides of the aisle, the two of them coming together with the same concern, which is
ultimately, what is this going to mean for the people AI cannot be controlled by oligarchs,
it has to be democratic, there has to be safeguards in place, and the benefits need to
come to all of us. But that kind of a call makes sense in a place like the United States, and it
makes sense in a place like Canada. Is China going to abide by any of the same rules?
Yeah, it's very difficult. And I understand what Trump is saying. And I think the real focus is
that, you know, part of what these companies have been saying is the first one
to reach what we call super intelligence, where AI can do everything as good as everybody, or even
better, is the one that's going to win everything. And that's the that's what the selling point has
been when they've been asking investors to invest. And so the concern, you know, especially from
Trump's perspective, I would imagine is that China will win that race, and will win the entire market.
China is very closely aligned with these companies were in the United States, it's a bit separate,
you have the government and you have big tech. But in China, it's more linked together. So
I think unless there's an international discussion around this between countries,
it's very unlikely that the decisions in Silicon Valley that they make are going to stop China,
especially because these things are coming to light now because China has caught up.
No longer is the US really dominant. These frontier AI solutions are not the fastest,
best solutions in the world anymore. And I think that's part of the reason that these
guys are starting to talk about it and raise their hands now.
Francis Sims, much appreciated.
Yeah, thanks for having me, David.
