Speaker A
let's look at AI because specifically the risk that could lead to the extinction of humans.
Experts warn AI poses extinction risks and societal challenges; urgent global regulation and ethical oversight are needed to mitigate harms.
Key Takeaways
- AI poses both immediate societal risks and potential existential threats that require urgent attention.
- Current AI development continues despite expert warnings, highlighting a gap between risk awareness and action.
- Effective regulation requires collaboration between governments and AI industry experts to keep pace with innovation.
- Balanced regulation should protect users and industry without stifling beneficial AI applications, especially in healthcare.
- Lessons from past technology regulation failures underscore the need for proactive safeguards against AI misuse.
Summary
- Top experts have signed a statement urging global prioritization of mitigating AI extinction risks alongside pandemics and nuclear war.
- The G7, EU, and US are actively discussing how to address AI challenges but have not halted AI development or investment.
- Stephanie Hare highlights current AI risks such as discrimination, misinformation, and election interference alongside long-term existential threats.
- Panelists emphasize AI's potential benefits, especially in healthcare, but caution about biases and regulatory lag.
- There is concern that AI developers continue building and profiting from AI despite warnings about risks.
- Experts call for collaboration between regulators and private sector AI experts to keep pace with rapid technological advances.
- Poorly designed or rushed regulation could harm both users and the industry, so thoughtful, informed policy is critical.
- Lessons from the internet's unregulated growth demonstrate the need for safeguards to prevent AI from becoming uncontrollable.
- The discussion stresses the importance of balancing immediate AI risks with long-term existential concerns.
- Overall, the video advocates urgent, coordinated global action to regulate AI ethically and effectively.
Full Transcript — Download SRT & Markdown
Speaker A
Many top experts have signed a statement warning of the risks of artificial intelligence.
Speaker A
And this is what that wording says: "Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war."
Speaker A
The G7 group of leading economies, the EU, US, well, they've all been holding meetings trying to work out how to tackle the challenges.
Speaker A
Well, I've been speaking to Stephanie Hare, a technology ethics researcher, about the current risks posed by AI.
Speaker B
They aren't talking about what they're doing to stop these risks from manifesting.
Speaker B
So, they are all still building this technology, they're not saying they're going to stop building it, they're building it.
Speaker B
And they're still seeking investment, and this investment isn't to the tunes of multiple billions of dollars.
Speaker B
So, that's not really a mitigation strategy, is it?
Speaker B
Without wishing to disrespect anyone on the list who are serious people and I listen to them.
Speaker B
There's a lot of people who aren't on that list, who are also very serious thinkers.
Speaker B
And who are warning of very different risks, not the sort of science fiction risks.
Speaker B
But the risks that are happening with AI right now.
Speaker B
And those are risks of discrimination.
Speaker B
Those are risks of misinformation and disinformation.
Speaker B
Those are risks to interference in our elections.
Speaker B
And it's interesting because if we don't talk about the risks that are happening right now.
Speaker B
Then they can carry on making money.
Speaker B
And they can carry on having us all think about things that may or may not happen in the future.
Speaker B
I would like to have us thinking about both.
Speaker B
Let's think about the existential long-term risks that are possible.
Speaker B
And let's think about what's happening right now and hurting people right now.
Speaker A
Well, the view there from Stephanie Hare.
Speaker A
Let's talk to our panel.
Speaker A
Aisha and Victoria are with us.
Speaker A
Is quite a warning, isn't it, Aisha, the mitigating the risk of extinction from AI should be a global priority, so says this warning.
Speaker A
Is this just scaremongering or should we be really concerned about this stuff?
Speaker C
I think we should take heed from what a lot of these experts are saying.
Speaker C
I think on the one hand, AI could provide some really important solutions to a lot of problems that we have in society.
Speaker C
Right across, particularly in the health sector, I was talking to a radiographer who was saying they're hoping to develop AI which can really look at cancer scans and things like that very, very quickly.
Speaker C
So you can see that there could be some huge advantages.
Speaker C
But, and there's always a but, I do think that there is a worry about how this AI is designed.
Speaker C
With inbuilt biases and also how do you regulate this stuff often, you know, policy makers.
Speaker C
Elected representatives are often very behind the curve when it comes to to keeping up with technology and how to regulate technology.
Speaker C
So I think we are right and politicians and regulators should heed the warnings, particularly from people who've been very involved in artificial intelligence from the the beginning.
Speaker C
And if they are sounding the sirens, you know, of of of warning.
Speaker C
Then I think, I think we do have to look at to and take it seriously.
Speaker A
Yeah, and that's the challenge, isn't it, Victoria?
Speaker A
Because, you know, that list of people warning against this is a who's who of people in the AI industry.
Speaker A
There's some very big names.
Speaker A
But I hope you could hear that clip we showed a little earlier from Stephanie Hare, and she's saying, look, the people that are warning against it are the very people who are still building it.
Speaker A
They're making money from it, they're getting research and funding for it, so if they were that worried, they should just stop, shouldn't they?
Speaker D
No, it's a very good point.
Speaker D
Good to be with you, Ben, and and Aisha.
Speaker D
And one thing that's been very striking to me is, as I touched on the AI world around 2019, 2020, when I was working for the US Department of Energy.
Speaker D
Which is one of our lead agencies on developing AI, and one of the things I was told then was that we were years and years away from this actually becoming a functional part of our society.
Speaker D
And it it is now suddenly happening in real time.
Speaker D
And so my concern is that we don't have a handle on this legally, ethically.
Speaker D
Uh, and that it is about to be enormously disruptive to our societies in ways we can only imagine.
Speaker D
And so agree with Aisha, we need to get in front of it now.
Speaker A
Yeah, and Aisha, you talked there about some of the the practical, very useful applications of this technology.
Speaker A
Particularly in health care, we talked just last week on the program about how it managed to whittle down a list of potential antibiotics to treat infection.
Speaker A
And it managed to save hours and hours of lab time to find the ones that could work.
Speaker A
But I wonder how we regulate, how do we separate the good from the bad?
Speaker C
Well, that is absolutely the the key question.
Speaker C
I think one thing that um government and regulators and and thinkers in this space should do is I think they should really collaborate with experts from the commercial world of artificial intelligence.
Speaker C
Because I think that with the best will of in the world, the expertise is not going to be found in government departments and with policy officials.
Speaker C
Because this technology and the the speed and the advancement just moves so quickly.
Speaker C
I think this is one area where regulators and policy experts really should bring in expertise from the private sector.
Speaker C
From these people who really are at the cutting edge because they're going to be the ones that that can help.
Speaker C
If not get ahead of this stuff, it's hard to keep at least try and keep up with what's going on.
Speaker A
Yeah, and Victoria, there is a danger, isn't there, that, you know, regulation done in a panic is not the best regulation.
Speaker A
It's not really got the best interests of both users and the industry at heart.
Speaker D
No, and I think we can take a sort of instructive lesson from what happened with the development of the of the internet.
Speaker D
And, you know, the the hope that that could be free and completely open and benefit everybody.
Speaker D
And it's it's has been obviously highly beneficial, but it there are also problems.
Speaker D
Because nobody figured out how to regulate uh, you know, the flow of information.
Speaker D
And how how this was going to be consumed.
Speaker D
So I I strongly agree that we're we need to figure out how you know, game these worst case scenarios of how this could become a dominant force that humans can no longer control.
Speaker D
And then put in safeguards, trip wires, warning bills, whatever you want to call it.
Speaker D
To ensure that that doesn't happen.
Topics:Artificial IntelligenceAI risksAI extinctionAI regulationTechnology ethicsAI misinformationAI biasAI in healthcareGlobal AI policyAI safety
