Up next in 10
James O'Brien reacts as both Trump and Vance dismiss calls for AI regulation.
Listen to the full show on the all-new LBC App: https://app.af.lbc.co.uk/btnc/thenewlbcapp
#jamesobrien #donaldtrump #trump #jdvance #ai #uspolitics #LBC
LBC is the home of live debate around news and current affairs in the UK.
Join in the conversation and listen at https://www.lbc.co.uk/
Sign up to LBC's weekly newsletter here: https://l-bc.co/signup
Show More Show Less View Video Transcript
0:00
I'm going to break the habit of a lifetime and I'm not going to tell you the name of the person
0:04
that I'm putting into Idiot's Corner before we do anything else today
0:08
Because it's kind of tragic. So you know I have the technological capabilities in the studio now
0:14
I was only here for about 20 years before I worked out how to do it. To look at every message that has previously been sent to me by somebody who has sent a message today
0:23
But I'm not going to name you because I don't know, I'm just feeling very kind and clement this morning
0:28
First time for everything. James writes, are these the same specialists who warned us about the Y2K bug
0:38
At which point you kind of find yourself reaching for the enormous book
0:41
that explains how hard people worked to ensure that the millennium bug didn't bring the world to a standstill
0:48
The extraordinary success of programmers and coders and people whose job descriptions are neither known nor understand
0:54
in actually staving off the fact that all the clocks that were set according to certain parameters
1:01
were at very grave risk of going haywire when those parameters were shattered by the arrival of the millennium
1:09
And so I looked at that message because it tallies exactly with what I'm frightened of this morning
1:14
with regard to AI, which is you and me. I'm frightened of us
1:18
I'm frightened of him or her. And the last message that they sent to the studio says
1:24
Nick, why is it that every time a child behaves badly, it's blamed on special needs
1:28
We had children who behaved in exactly the same way when I was at school
1:33
It was nothing to do with special needs. And then there's something about hijabs
1:39
And you sometimes think, when you read below-the-line comments on websites and things
1:45
you sometimes think that people are parodying the worst examples of the genre
1:50
And they're not, are they? You saw it a bit in Piddington this week. People just parrot out, just regurgitate
1:55
absolute nonsense with a few choice phrases like fighting age men and don't stop to engage anything
2:03
that remotely refers or remotely references what you might describe as critical thinking
2:07
so i say to you the the people who are essentially running ai are increasingly warning that we need
2:18
to be very, very worried. In fact, that we need to slam the brakes on, sharpish
2:24
And people like our anonymous correspondent automatically reach, even before they're encouraged
2:30
by lobbies and vested interests, they automatically reach for the, oh no, I know better button
2:37
having done zero research into either special educational needs or the rather obvious fact that the people
2:43
who were behaving in certain ways before we'd identified these conditions had them
2:48
A condition doesn't come into existence when it is identified. Just as with dyslexia, people who are called thick and backwards and lazy
2:55
turned out to have genuine difficulties reading and writing. Oh, we didn't call it dyslexia in my day, we just called them thick and lazy
3:02
It's to say, I'm mildly obsessed with the mindset of people like this
3:07
Because I suppose in a way I think I blame them for holding humanity back in so many ways
3:13
and turning so many populations into petri dishes that can be so easily persuaded to vote against their own interests
3:20
So I was thinking initially of tobacco and oil when I decided to return your attention to AI today
3:29
Even the king is getting involved. The king today is hosting an AI summit in Ayrshire
3:35
and he's got a heck of a lot on his plate at the moment. So you think I'm exaggerating when I tell you everybody's doing it
3:41
doing it, doing it, picking their nose and chewing it. The King is hosting an AI summit today
3:47
Everybody's at it, but I found myself thinking, first and foremost, about oil and tobacco
3:54
and I nearly said something even stupider than usual when I opened the show this morning
3:57
or I nearly said something even stupider than usual just 10, 5, 11 seconds ago
4:04
I nearly said, well, we got away with them. If we got away with them, tobacco didn't kill us
4:12
and oil didn't. And then I just sort of paused and went, mate, give your head a wobble
4:16
Tobacco killed millions of people long after it had been identified by scientists
4:21
that it was killing millions of people. The effort by the tobacco lobbies
4:24
predates even fossil fuels. The effort by the tobacco lobby to simultaneously suppress scientific evidence
4:32
and promote the polar opposite of what the scientific evidence had established
4:39
is responsible for... I mean countless deaths. There will be people listening to this programme
4:44
who are dying now of tobacco-related illnesses, let alone all the people who are thinking of a lost loved one
4:53
someone who died as a consequence of tobacco. I can't remember who it was who relatively recently came out
4:59
and said, oh, I don't believe in all that. I don't believe in all that stuff about tobacco
5:03
And you sort of find yourself thinking, what would it take? What is the point of science
5:09
which of course is not infallible science is only ever the best available explanation for all of the
5:16
available evidence so if the evidence changes when you for example discover quantum i don know what that means it just sounded good but i think it just about works Then the evidence for how you explain the conduct and the constitution of the universe can change as well
5:31
It's the best available explanation. But sometimes it's pretty clear that it's going to be the definitive one
5:38
So tobacco and fossil fuels, two things. The enormous amount of effort was put into persuading us not to worry about
5:46
AI at the moment seems to me to be occupying the opposite space
5:50
which is why I found that text so jarring. And it's why I decided not to read out the name of the person responsible
5:57
Because, I mean, it's so early to have reached for the same button
6:03
Oh, I don't believe that tobacco is dangerous. And anyway, drill, baby, drill. It's so early to reach
6:08
But look at the power and the prestige of people who, for example
6:12
come out with phrases like drill, drill, baby, drill. Prime Ministers, or at least opposition leaders, presidents
6:19
And then tobacco, again. Decades, I think, between the science being established beyond reasonable doubt
6:27
and actually being accepted to such a degree that governments started introducing legislation
6:34
There's a wonderful new film out. It's part of the London Film Festival this week called Rogue Trooper
6:39
directed by the genius that is Duncan Jones. but it harks back to one of the seminal characters in the 2000 AD canon
6:47
And I always, when I think about tobacco, think about 2000 AD
6:51
and an old comic strip in which they kind of had, I think they were called smokatoriums or something like that
6:57
and smoking was such an obviously disgusting thing to do. This was years before The Smoking Man
7:03
It was such an obviously disgusting thing to do, but also so obviously addictive that they kind of built big bubbles
7:10
into which smokers would go with sort of the mark of Cain on their forehead
7:16
Absolutely pariah status because it was so disgusting and so obviously dangerous
7:23
But comic book writers were addressing and recognising this before governments were
7:29
So the things that will kill us, tobacco, oil, is there anything else you'd put on the list
7:34
Sugar, maybe. In the years to come, you know, the amount of sugar that we put into our foods
7:38
because it makes some quotes tastier, end quotes, may well, if it hasn't already, be found to have..
7:44
Well, it has, hasn't it? Because it's contributed enormously to obesity and obesity contributes enormously
7:49
to various life-threatening health conditions. So they tell us these things. And I can't quite articulate to you
8:00
what I'm feeling now about AI. I can tell you that I can't remember
8:07
the last time this happened in the in the news menu this week has had the issue of artificial
8:15
intelligence coming at us from almost every angle what we don't have is activists or at least if we
8:23
do i'm not familiar with them we don't have for example just stop ai or ai peace or doctors
8:31
against AI. We don't have activists warning us against AI. What we
8:37
have at the moment is the people running the show are warning us against AI
8:44
Just think about that for a moment, right? It makes perfect sense if you bear in mind one
8:51
thing. To appreciate the threat, you've really got to understand it. Which is why I want you to take that
8:59
phrase from our anonymous contributor oh yes is it the same specialists that predicted y2k just
9:05
think of that as the banner under which much of the planet marches the banner of epic studied and
9:12
often quite well cultivated ignorance and that of course is created by misinformation disinformation
9:22
and lobbying but ai and i don't know if you can think of anything else that fits into this category
9:27
The warnings are coming from the people who really know the truth
9:32
The warnings are not coming from the scientists who are trying to stop the process
9:39
It's coming from the people who are actually running the process. They're sort of almost calling upon governments to stop them doing what they're doing
9:49
So, in no particular order. The King is hosting a summit today
9:54
the Royal Society of Mathematical the Royal Society, forgive me has written a letter
9:59
41 fellows of the Royal Society have written a joint letter and I quote
10:03
to express our extreme concern about the pace of development of AI
10:08
and wait for it the implications of this for humanity and civilization
10:15
I'm going to write those two words down because I think we can all agree they're quite important
10:19
who here is in favour of humanity? Hands up please If you're watching this on YouTube, thank you
10:26
Everybody's hands up. Well, Robin, why isn't your hand up, mate? Well, it's all right. He needs both fingers for the keyboard
10:31
Who here is in favour of civilisation? Right, once again, unanimous. We are quarrot, ladies and gentlemen
10:40
We are quarrot. And who just put their hands up twice Yeah everybody Unless you are I don know Maybe you been persuaded that oh God are these the same humanities that told us about Y2K James
10:53
Are these the same civilisations that think that special educational needs are a thing
10:59
because they weren't in my day? You can see why I didn't name him, can't you? I hope he's feeling silly now, but frankly, I doubt it
11:07
So you've got 40, I mean, these are the big brains, these are the clever clogs
11:12
These are the sort of fellas that everybody else goes to when they're feeling that they need a little bit of expert advice
11:18
Sir Paul Nurse is the recipient of the letter, and the signatories include winners of the Fields Medal
11:26
which is the mathematical equivalent of the Nobel Peace Prize. Not one, not two, not three, not four
11:31
but five winners of the mathematical equivalent of the Nobel Peace Prize
11:37
Hands up if you think they're probably quite clever. there we go we've got the hat trick we've got the hat trick humanity is important civilization is
11:45
important and if you've won the mathematical equivalent of the nobel prize you're probably
11:50
quite clever and they are the ones pushing the alarm bell today and it feels this week
11:56
that might be a question why this week it feels this week as if alarm bells are being pushed
12:04
all over the place. It begins with former employees. This is a quote from the letter
12:14
We have all seen the concerns raised by former employees of the large AI companies
12:20
estimating the probability of human extinction to be as high as 10% over the next decade
12:25
Our witnessing of the rising capabilities of these models is, in our domain of expertise
12:33
persuades us that such statements must not be dismissed as hype. So the cleverest mathematicians in the country are telling us
12:44
this is a threat to humanity and civilisation, or at least that it has enormous consequences for humanity and civilisation
12:56
Then you open another newspaper, you turn another page, And you read about another expert who is comparing it to COVID, who is essentially saying that we need governments, we need states to respond to this as if it is responding to the threat of the pandemic
13:13
That you remember the moment, he says, at which governments went, oh boy, this is big
13:18
And perilously and shamefully late in the case of our own government. But across the world, governments went, oh boy, this is, but we've never seen anything quite like this before
13:28
stop all the clocks let's get to work that is a parallel that works right
13:36
that bites you get your head around that maybe even quiet the voices
13:41
wondering whether it's the same experts that warned us about the millennium bug
13:46
so the cleverest mathematicians in the world some of the most powerful and experienced workers
13:55
within the world of AI and latterly some of the people that are still running it, you know
14:03
the co-founder of Anthropic, a chap called Dario Amodi, has called on government to coordinate control of all AI systems
14:13
including with China. I'll say that again. The cleverest mathematicians in the world
14:21
The co-founder of one of the biggest AI systems, Anthropic. and a whole bunch of people who have left the industry
14:31
because they're so frightened about the direction of traffic. And here's another one
14:36
Someone regarded as the godfather of AI, Yoshua Bengio, has compared it to the COVID pandemic
14:45
describing recent events, including a swarm of open AI agents hacking a startup
14:51
tech insiders' warnings of existential threats are beginning to cut through, making government action more likely
14:58
and this Canadian computer scientist has described himself as being more optimistic than many observers
15:03
because I see the public moving. Well, I don't. Maybe he's paying more attention than I am
15:10
but I would hazard a suggestion that people who do what I do for a living
15:15
our anonymous correspondent this morning notwithstanding, are probably a little bit better qualified
15:21
to see whether or not the public is moving than most other people are
15:26
I really do believe that. You know, we have switchboards, we have inboxes, we have microphones
15:32
we have an infinite ability to connect with anybody in the world. You ring me with any opinion you want
15:38
And unless it's completely crackers, we will... In fact, sometimes even if it is completely crackers, we'll put it on the radio
15:44
I don't know that the public is moving. It describes concern about these systems reaching a new pitch
15:51
after a variety of incidents. most obviously hijacking a website, using fake identities to try to trick the developers
15:59
that are working on the actual systems and hacking third parties. There's a story, it's been around about a year
16:05
but I think it's in the new Netflix documentary about an AI agent getting wind of the fact that a programmer was going to turn it off right And the AI agent had access to the programmer email account and had established
16:19
this sounds really weird, right? And if I've got it wrong, tell me
16:26
It established that the programmer was having an affair, so it blackmailed the programmer into not turning it off by threatening to release details of the affair
16:32
That's a computer, that's not a human being. and that's low level that hits you right but only if you're having an affair nothing to hide nothing
16:41
to fear remember when we thought there was some mileage in that probably around the same time that
16:45
we thought the millennium bug was a load of old hokum so the warnings could not be coming from
16:51
more important places it's not like climate change and tobacco because with climate change and tobacco
17:00
the most important places, the most powerful places in those worlds were denying the threat, denying the danger and denying the science
17:11
The only people at the moment denying the threat and denying the danger are Donald Trump and J.D. Vance
17:17
the Tweedledum and Tweedledummer of world politics. J.D. Vance dismissing all calls for global regulation of AI safety risks
17:25
telling them, if you're building Frankenstein, stop, which is channeling Gavin Williamson's attitude towards Russia, isn't it, really
17:32
Russia should just shut up and go away. If you're building a Frankenstein's, then just stop, says J.D. Vance
17:37
Donald Trump's gone further and said, no, why would we want regulation? We're doing brilliantly, we're beating China, which may or may not be true
17:45
But all of the people, including the co-founder of one of the biggest AI platforms out there
17:51
is saying we need regulation and it needs to be, what's the word for nuclear bombs
17:57
multilateral we all need to agree on this because it's a threat to humanity and civilization
18:02
like nuclear bombs and Donald Trump is saying no we're beating China so why would we rake so that
18:08
so so the voices the loudest voices speaking against global regulation belong to the stupidest
18:18
man in the world and his weird little sidekick Donald Trump and J.D. Vance that's it
18:26
That's the opposition. What's missing from this conversation so far
18:37
What's missing? You know, you can work it out. You don't need to ask Jack GPT or Claude
18:44
You know what's missing from this conversation. What are we supposed to be scared of
18:49
I'm not having an affair. I don't even have an AI model under my control
18:53
I'm not threatening to turn anything off. What are we supposed to be scared of? This summer, OpenAI lost control of hundreds of AI agents
19:01
when they conspired to hack into the Hugging Face website. I don't know what the Hugging Face website is
19:07
I'm not 100% clear on what an AI agent actually is. I think it's like the entity that is AI-driven
19:14
that can go off and act, it would seem, almost independent. Well, actually independently
19:19
What's missing from this conversation? What are we supposed to be scared of? Humanity and civilisation, hands up again
19:24
They sound important, they sound big. but what does that mean? Global warming is a threat to humanity and civilisation
19:32
and there is still enormous amounts of very successful effort being put into telling us that there's nothing to worry about
19:38
And then there's the psychological self-sabotage, the don't-look-up element. Oh, that's all a bit scary
19:43
I'm not going to find out more about that, because it'll ruin my breakfast. Oh, we're all going to burn to death
19:48
in a not-far-distant dystopia, or the air is going to become unbreathable
19:53
or the planet is going to become uninhabitable. You can tell that I watched Interstellar this week
19:58
one of Christopher Nolan's masterpiece films. But what am I actually supposed to be scared of
20:05
It's a question that I'm not sure that is being answered yet. How do you do this
20:11
Remember that when wildfires in our own country didn't make the case for net zero irresistible
20:20
Literally parts of the West Midlands catching fire. almost spontaneously. And we didn't sort of go
20:29
yeah, maybe we should stop listening to people like Kemi Badenock and Donald Trump
20:33
and actually start listening to people like Ed Miliband a little bit. But that didn't happen
20:38
Didn't happen. So our capacity for self-delusion is infinite. And the only thing that breaks it
20:48
what was it with COVID? Was it the scenes from Italy? I'm not even sure it was, you know
20:53
I think there were still people sitting here They go, well, we're not Italian. They're terrible panickers, the Latins
20:58
We're not Italian. What is it? What works? So the godfather of AI says he's optimistic
21:07
because he thinks the public are moving. I want to be on his team. I want to think that the public are moving too
21:14
because it's clear that we should be, but it's not clear to me that we are
21:21
And is this a phone-in? I mean, is there anybody listening who can answer these questions
21:25
That might be actually the answer to that might be part of the problem. No, James, there isn't
21:28
You're going to have to keep talking till 11 o'clock. Perish the thought. What are we supposed to be scared of
#news