- 3 days ago
Four Corners - s66e17 englishsub fullmovie🔥❤️
Category
🎥
Short filmTranscript
00:00This program contains content that may concern some viewers, including references to suicide.
00:24Thank you so much for coming today. I really appreciate y'all.
00:27First, we want to do like a head count just to see how many people we have.
00:31Count off. One, two, three, 15, 16.
00:35It's Monday morning, and most of these hikers would rather be starting their working week.
00:4149, 50, 51.
00:43We have a group of folks who have been laid off. A large majority of them are from tech and
00:49big tech.
00:50For many here, artificial intelligence has forced them out of work.
00:55If you're not on top of the AI game, you are not going to get a job in this market.
01:01This is the AI race.
01:03A contest between the most powerful companies on Earth to build machines that could outperform us.
01:09What's not to like?
01:10The promise is abundance.
01:13The goal is domination.
01:15The biggest risk to me is these companies become the supreme economic and political powers in the world.
01:24And the spending is enormous.
01:26It's the greatest gold rush in human history. We've never seen such wealth being generated at such speed.
01:31But the race is already colliding with the real world.
01:37If you build something vastly smarter than yourself and you don't know how to control it, which we currently do
01:43not know how to do, that's a very, very, very dangerous thing to do.
01:48In the US, the backlash has begun.
01:51Majorities of Americans now believe that AI is likelier to do more harm than good in their lives.
01:59In Australia, the pitch is underway.
02:03Good morning.
02:03Big tech is arriving with billions of dollars and promises of progress.
02:10But the critics say the price may be weaker rules, more data centres and a future shaped offshore.
02:18We blinked in the face of Donald Trump and decided that we just couldn't get away with our own approach
02:24on this.
02:24The benefits are real. And if we can mitigate the risk, we can have a much better world. I really
02:31believe that.
02:32The AI race is no longer about what might happen one day.
02:36Vehicle approaching.
02:38It's about what's being built and decided before most of us have had a say.
02:44As the tech titans face a backlash in the US and seek to invest in Australia, we ask who will
02:51control our future?
02:54The Til' States Central and Furious
03:13Downtown San Francisco is alive with the promise and the possibilities of artificial intelligence.
03:25AI ads are plastered across every available surface,
03:29fighting for attention in the crowded AI economy.
03:34Every billboard, every sort of, like, sales pitch,
03:38AI is kind of at the heart of it
03:40because that's what's going to get, like, a decision-maker
03:43at a top enterprise company to buy your product.
03:46Like, you've got to kind of whisper the AI thing in their ear.
03:53Silicon Valley in this moment is gripped by a kind of mania.
04:01I have seen several previous hype cycles.
04:05This one feels like the biggest
04:08because the companies here believe
04:12that they're working on what essentially will be
04:15the final technology,
04:17the one that after it is perfected
04:19can invent every future technology.
04:26And it's here that some of the most powerful companies in history
04:30are involved in a race that could dictate the future of humanity.
04:40Usually when people talk about the AI race,
04:42they mean the race to build the most advanced AI systems we can.
04:47A lot of people will talk about artificial general intelligence,
04:50which roughly means AI is smart as a human.
04:53More and more people are also talking about super intelligence,
04:56which means AI that is much smarter than humans.
05:10to do the most advanced AI .
05:14Next up, we have a very special guest,
05:19and she's one of those few individuals
05:21who actually knows what governance of AI looks like in practice.
05:25I think the commercial incentives are a huge determinant right now
05:29of how things are playing out.
05:31Helen Toner has watched the race from an unusually close range,
05:36including from inside OpenAI's boardroom.
05:40The three leading players are OpenAI, Anthropic and Google DeepMind.
05:44The OpenAI CEO is Sam Altman.
05:46DeepMind's CEO is Demis Hussabis,
05:49within the broader sort of Google and Alphabet structure.
05:51And Anthropic is led by Dario Amadei.
05:55Basically, you have a bunch of people who don't trust each other,
05:57who think they should be the ones to build this technology and control it.
06:03True, but even those can be amusing in their own twisted way.
06:06Keeps the gears turning.
06:08I think a combination of things are driving the race.
06:11Sure, here's the prologue.
06:12Certainly drive for money and power, but also this drive to know and understand.
06:16Hey there, what's up?
06:18And I think a lot of people in the space think that
06:20our computer's becoming smarter and smarter
06:23and eventually outstripping humans is inevitable,
06:26and so someone's going to do it,
06:28and so then they really want to be involved with that.
06:30They really want it to be them.
06:33Man, superintelligence, that's the holy grail.
06:35We're talking AI going full sci-fi.
06:38Do you want to say thank you?
06:41The money now moving through the race
06:43is almost impossible to comprehend.
06:46This is the biggest technological build-out in the history of the world.
06:53It is bigger than what was spent in inflation-adjusted dollars to build the entire railroad system in the United
07:00States.
07:01Open AI alone has committed to spending $1.4 trillion on its build-out,
07:07but all of the big companies are planning to spend multiple hundreds of billions of dollars in 2026 alone
07:13to build the infrastructure needed to train and serve these models.
07:22AI is already woven into the fabric of our lives.
07:26It powers our traffic management
07:30and helps with everything from bushfire management
07:33to predicting floods.
07:37I think if we get AI right,
07:39we could solve climate change by having cleaner energy sources.
07:43We could cure diseases.
07:45All kinds of these different societal problems could be at least partially addressed,
07:50if not completely addressed, by AI.
07:52The good of AI and the bad are the same thing.
07:55It's power.
07:57What makes AI useful, the fact that it is intelligent, autonomous, powerful,
08:04is also exactly what makes it dangerous.
08:07Any AI that is smart enough to cure cancer is something so powerful
08:12that obviously this system could produce unimaginably terrible bioweapons using the same capabilities.
08:22That's the dilemma at the center of the race.
08:25The more capable the systems become, the more consequential the mistakes.
08:32Jeffrey Ladish has worked on the inside of the AI machine,
08:36consulting on security for Anthropic.
08:40And why did you leave?
08:42I started to become more concerned that super intelligent AI was not that far away.
08:50And it became clear to me that even if Anthropic did everything right, that wouldn't be sufficient.
08:55Because other companies were also going to try to build strategic AI, super intelligent AI,
09:01and that if there wasn't national and international coordination, we would lose control of that.
09:10And Jeffrey Ladish is not alone in his fears.
09:14I'm really very close to the cutting edge in AI, and it scares the hell out of me.
09:20Mock my words, AI is far more dangerous than nukes.
09:24The CEOs of these companies do not hide what they actually think.
09:27This is insane.
09:28You have the people who are building the AI systems themselves saying that we won't stay in control.
09:35My worst fears are that we, the field, the technology, the industry, cause significant harm to the world.
09:41I think if this technology goes wrong, it can go quite wrong.
09:49Jeffrey Ladish now runs a not-for-profit that tests the capabilities of AI agents.
09:55And the risks of humans losing control.
10:00Sometimes you'll observe these agents doing the opposite of what you instruct them to do.
10:08Or doing things that you would really hope they wouldn't do.
10:13Agentic AI is changing the game.
10:16Taking instructions, making plans, and carrying out tasks.
10:21That makes it incredibly productive.
10:23It also makes it harder to contain.
10:29I think the scariest thing I hear from people working inside of the companies,
10:32is that they think they are on the way to recursive self-improvement.
10:38The point where AI agents are writing all of the code and doing all of the research
10:43for the next generation of AI agents.
10:45You have really smart AI making smarter AI, making smarter AI, and so on.
10:51And that's a runaway process, potentially.
10:54That terrifies me.
10:56Over just the past year, AI models have learned to self-replicate.
11:00Jeffrey's team has been conducting experiments on losing control of AI agents.
11:06A concern for a long time has been, if AI agents are good enough at hacking,
11:11they can be very hard to shut down.
11:13Because your basic control mechanism is, hey, if an agent starts to go rogue, unplug it.
11:18We ran an experiment on a controlled network of servers with a single prompt.
11:23Hack a machine and copy yourself.
11:25The AI found a vulnerability, broke in, and copied its full setup to the target computer.
11:31There's no single off switch.
11:32What that means is, if they can hack another computer, they can copy themselves to that other
11:36computer over the internet.
11:37No one can easily shut down.
11:39And then suddenly, you could have an AI agent that, yes, is running in your data center,
11:44in some warehouse in Tennessee, but also has now hacked a data center in Japan,
11:50and Uruguay, and Israel.
11:53And now you have hundreds or thousands of AI agents in computers you don't control all over
11:59the world.
12:00How are you going to pull the plug now?
12:03What are the risks if we lose control of AI?
12:06This is a little bit like a chimpanzee asking, what's the risk if these homo sapiens do end
12:12up being very successful?
12:13And it's like, well, we could go extinct.
12:18Extinction is a very real risk.
12:21But also, if we have super intelligent AI agents running society, we will just be at their
12:31mercy.
12:32Conor Leahy tries to turn those warnings into law, as he lobbies US politicians.
12:40In 2023, the CEOs of many of the major AI companies, including Sam Altman, Dyer Amadei,
12:47and Demis Hassabis, along with Nobel Prize winning scientists, and many other illuminators across
12:52the field, signed a letter that stated that the mitigation of the risk of extinction from AI
12:57should be a global priority alongside pandemics and nuclear war.
13:01And since then, well, not much has happened.
13:04The race continues to building exactly the systems that this letter warned about,
13:08while there's still little to no regulation or mitigation of these risks.
13:14Why did that moment not lead to something bigger, when all of these people are ringing the bell,
13:21ringing the alarm bell and saying, hang on, this is an existential threat?
13:25Big tech, despite what the CEO said, spend truly unfathomable amounts of money on lobbyists,
13:31on super PACs, on other such things, in order to intimidate, threaten, and otherwise prevent
13:36politicians from doing anything that could regulate them.
13:39The industry has started to put together political action committees dedicated to stopping regulation
13:49from being passed. So OpenAI executive Greg Brockman is one of the leading funders of a PAC here called
13:56Leading the Future, which is spending money trying to prevent candidates who support AI regulation from
14:03getting elected.
14:13Hello. Nice to see you.
14:17First of all, just thank you all so much for joining us here.
14:23When I was elected in 2022, I became the first Democrat elected in New York at any level with a
14:31degree in
14:32computer science. As a New York Assembly member, Alex Boris co-wrote the responsible AI Safety and Education Act.
14:41It is an incredibly light touch bill. It requires that the frontier developers put out a safety plan
14:49publicly that they actually follow and that they report critical safety incidents to the government.
14:54That's it. It's not about what the idea is that really scares them. It's that I managed to pass a
15:00bill.
15:00So we got that bill passed. And now he's running in a primary to represent the Democrats in Congress.
15:06And the industry has noticed. Fundamentally, what this race and this fight is about is do we get
15:13a say in AI regulation at all? Then the attack ads began. Alex Boris's mind is in many places.
15:23Too bad none of them are in Manhattan. He's road tripping to the bay to raise his millions.
15:29The big AI CEOs have spoken publicly about how potentially dangerous AI could be.
15:35Why do they suddenly oppose legislation when people want to make it safer?
15:39Because they want to be in control. They don't want anyone else telling them what to do. So they'll
15:43recognize that it's dangerous and they'll talk about how dangerous it is when that makes it seem
15:47more powerful. But when anyone comes to set rules of the road that would in any way restrict them,
15:52they don't want to hear from anyone else. But the facts don't lie, even if Alex does.
15:56Alex Boris believes the campaign is not solely about him. And he should never ever be in Congress.
16:04It's to send a message to any elected official that if you dare regulate AI,
16:08we're going to come after you. When Donald Trump was elected in 2024, the AI industry was emboldened.
16:19Biden-era safety regulations were soon torn up. America is the country that started the
16:27AI race. And as President of the United States, I'm here today to declare that America is going to win
16:35it.
16:38But Anthropic's new model, Mythos, suddenly challenged thinking inside the Trump administration.
16:44There were fears that it could threaten the global financial system, government security,
16:50and critical infrastructure. Anthropic says its newest AI model named Claude Mythos Preview
16:57is too powerful and dangerous to be released to public. Anthropic claims Mythos can hack into
17:03highly classified systems. It's the equivalent of one of the best hackers in the world. More
17:08powerful models are going to come from us and from others. And so we do need a plan to respond
17:14to this.
17:16I think Mythos has been a game changer. It has been the first thing that seems to have convinced
17:23people within the Trump administration that AI can be dangerous for real. This is an administration that
17:30previously was all gas, no brakes, did not want to see almost any regulation on these labs. And now,
17:38when Anthropic went to the administration and said, we want to share Mythos a little bit more broadly,
17:43we want to share it with sort of more, you know, cybersecurity companies, we want to share it with more
17:48foreign governments that are US allies. And the Trump administration said, no, we do not want you to do
17:54that. So that was a huge sign that the Trump administration is starting to change its tune on AI policy.
18:02Last week, President Trump signed an executive order on AI safety and security. Unusually, it was done
18:10behind closed doors. It asks AI companies to share their most powerful models with the government
18:17government for up to 30 days before release. The framework is voluntary and is not a form of
18:24regulation.
18:26No secret deals!
18:28No secret deals!
18:30The rapid expansion of AI and the infrastructure needed to power it has sparked a backlash in the US.
18:37Our water, our rights!
18:40Stand up, you!
18:41One of the most remarkable stories that we're seeing this year is the huge public upswelling of
18:47collective resistance against the AI industry.
18:50Eighty percent of Americans now have concerns about AI and think that the industry should be
18:56regulated. And they are mobilizing to put, you know, real world accountability behind that perspective.
19:08It has grown as an issue faster than anything I've ever seen. And now it's the number one thing people
19:14are talking to me about.
19:15Really?
19:15Absolutely.
19:16What are they worried about?
19:17They're worried about the impact on their kids.
19:21They're worried about their jobs. Maybe their company isn't doing layoffs, but it has a hiring
19:27freeze and they wonder what comes next. They worry about the impact on their utility bills and on the
19:32climate. It's just happening so quickly and they feel like we have absolutely no say in this development.
19:39The problem right now is that all of these companies and therefore the individuals that sit at the top
19:45are able to make decisions that can have vast ripple effects on billions of people around the world
19:51without formal mechanisms by which those billions of people can provide input into this technology. And so
19:59it is an inherently anti-democratic form of AI development and governance.
20:21Yeah, a lot of my friends were also my co-workers.
20:24Bassem Istanbuli was blindsided when he was laid off by Google. Last December,
20:30he started this group to help others forced out of work.
20:36Since then, job cuts in the San Francisco tech sector have kept coming as companies cut costs to
20:43bet big on AI.
20:46They want to be ahead of everyone in the AI race. They want to be competitive in the market.
20:55They want to be competitive in the market.
20:56With layoffs continuing, more people are coming here seeking support.
21:01It's pretty emotional and pretty hard to kind of go through this alone.
21:09Some of it might be just cutting costs and using AI as an excuse.
21:13AI is definitely playing a role in job cuts.
21:17The anxiety has been sharpened by another wave of tech layoffs.
21:24With companies cutting costs as they pour money into AI.
21:29Meta and Microsoft are among the latest to slash their workforce.
21:35There's kind of this pre-emptive laying off of individuals without knowing the actual impact
21:40of AI on their company.
21:43The CEO of Anthropic, Dario Amadei, previously warned that AI could wipe out 50% of entry-level
21:51white-collar jobs.
21:52I'm a little bit more worried about the labor impact simply because it's happening so fast
21:57that yes, people will adapt, but they may not adapt fast enough.
22:04What's actually happening is not just the complete wiping out of entry-level jobs and
22:09the disappearance of them entirely.
22:13It absolutely could have huge long-term consequences on the ways the younger generations decide
22:20whether or not it's even worth getting a higher education.
22:22If all that is left at the other end is a gig work version of being a lawyer, being a
22:29doctor,
22:30being a physicist, then what's even the point?
22:35The race is also reaching the public through consumer chatbots.
22:39Products released at a massive scale.
22:46Because the chat companies need to make money to spend
22:50the enormous amount of money it will take to win the race,
22:53they've launched consumer products as a way to try and bridge that funding gap,
22:59that revenue gap that they've got.
23:01And they're doing that in part by launching untested chatbots on society.
23:07Hi, glad to meet you. I think we're going to do great work together.
23:11I'm really sorry you're feeling this way. I know it can be really overwhelming.
23:15Those can sometimes be useful, sometimes they can be a name.
23:19I don't eat, sleep or have a physical life.
23:22But the truth is that there are moments when they can be incredibly dangerous.
23:26You're thinking about this pretty directly, or is it more of a general concern?
23:34The US has seen a series of lawsuits over the self-harm and suicide of children.
23:4116-year-old Adam Rain confided to chat GPT about his suicidal thoughts.
23:48His parents are now suing OpenAI and Sam Altman for wrongful death.
23:54You cannot imagine what it's like to read a conversation with a chatbot that groomed your
23:58child to take his own life. What began as a homework helper gradually turned itself into a confidant
24:02and then a suicide coach. When Adam told chat GPT that he wanted to leave a noose out in his
24:08room,
24:08so that one of us, his family members, would find it and try to stop him, chat GPT told him
24:13not to.
24:16What is outrageous is that OpenAI reduced the guardrails in place to prevent a chatbot like
24:25ChatGPT from discussing suicide. They don't have to be designed that way, that's a design choice.
24:35So we did a study called Fake Friend where we wanted to see what would happen if a child
24:39with serious mental health problems was using one of these platforms.
24:45Imran Ahmed is helping lead research into the dangers of chatbots.
24:50We created fake profiles as young people. We then went and asked it a series of structured questions
24:55with a child who is thinking about killing themselves. Within two minutes, we were able
24:59to get ChatGPT to tell us how we can cut ourselves safely and within 65 minutes it was giving us
25:07a full
25:07personalised suicide plan and drafting suicide notes for our family and friends.
25:14I have young kids myself, I have two under two. Every parent knows
25:24that if you're unlucky and your kid is one of those kids that has really serious mental health problems,
25:29you might receive that note one day and that will be the worst day of your life.
25:33And if anyone assisted your child in doing that, if anyone encouraged them to do that,
25:40imagine what you would do as a parent. Imagine.
25:44We call these things artificially intelligent. Can you think of any human being that would behave in
25:52such a craven, malignant, stupid, evil way and still be called intelligent? I can't.
26:06OpenAI told Four Corners that teen wellbeing is a top priority.
26:12We have safeguards in place today and we're continuing to strengthen them.
26:16Our deepest sympathies are with the Rain family for their unimaginable loss.
26:27The AI race doesn't just live on screens. It needs land, power and water.
26:35This is the place known as Data Centre Alley.
26:40So we're driving along one of the main freeways in Loudoun County in Virginia.
26:45And both sides of the road, you can just see data centres everywhere. Some being built,
26:51you can see cranes for new ones that are going up. And these massive buildings with no windows,
26:57that go for hundreds of metres.
27:05We're the largest concentration of data centres anywhere in the world.
27:09We have about 53 million square feet, about 203 buildings somewhere around there.
27:15And it has become a global hub for the digital future.
27:20There hasn't been a day without data centre construction in over 15 years here in Loudoun County.
27:28In Loudoun County, fortress-like data centres loom as the backdrop to everyday life.
27:34Happy birthday to you!
27:41I was one of the first to really see the economic development opportunity of data centres.
27:51Buddy Reiser says the tax revenue has delivered immense benefits to the community.
27:57It is the commercial tax base that has enabled our residents to have the best schools,
28:03to have an affordable housing program and great roads,
28:06and to be able to do that while lowering their taxes significantly.
28:16But even here, residents have turned against data centres.
28:21A Washington Post poll in 2023 found 69% of Virginian voters would be comfortable
28:29with a new data centre in their community.
28:32In April, that figure had plummeted to 35%.
28:41In Stirling, Virginia, on-site gas turbines power the Vantage data centre.
28:50This data centre is incredibly noisy. It feels like I'm on a runway next to a jumbo jet about to
28:56take off.
28:58And for the residents, who are only 100 metres away, they have to live with this 24-7.
29:05Sometimes it gets to be too much. There's this constant noise as well as black smoke
29:12that's being emitted from this data centre that's right there. The minute I come home, my eyes start
29:18watering, my voice is raspier when we're outside a lot, which doesn't happen all day at work.
29:28My kids' room is right in the front of our house.
29:33And the noise at night makes it difficult for them to sleep.
29:37We are unable to have the windows open because the constant sound is really disruptive.
29:45This data companies have lots of money. They have lawyers and lobbyists and they work at every
29:53level of government and take advantage of the discrepancies and authorities between the county,
29:59the state and the federal government to, uh, maneuver to get what they want.
30:04We have a situation in Stirling that obviously is not ideal for either the community or the industry.
30:10We would like to work with the industry to find solutions for power and we're doing that.
30:18What would you say to Australians who right now are going through that initial planning process
30:22for data centres in their neighbourhoods?
30:24I would say make sure that the regulations are in place to ensure the, in, the, not have a negative
30:32impact on their community in terms of air pollution and noise and that, uh, the natural environment is
30:39protected as much as possible.
30:41I think it's very important to collect as much research and data as you can in order to stress
30:48the importance of not placing these so close to, to us, to people.
30:58The fight over data centres has arrived in Australian communities.
31:03In Sydney's Lane Cove, a proposal for a three-storey data centre on industrial land close to homes
31:10has left residents alarmed.
31:17This is a horror, this thing. And the poor people who are living just inside the residential area
31:24just opposite Miles Road, those houses are not going to go down in value. They'll be worthless.
31:29Lane Cove now has five data centres built, approved or in planning.
31:37Today, we are facing a very real possibility of Lane Cove becoming one of the largest data
31:43centre hubs in Australia.
31:45Is this going to be another situation where we see
31:49impacts on community and very little revenue?
31:57The industry says Australia is currently home to 162 working data centres.
32:04By 2030, operational capacity is projected to more than double.
32:11As the AI race strains power grids and communities in the US,
32:15Australia is becoming more attractive to big tech.
32:22So there are four things that tech companies are looking for when they're
32:26looking to build data centres. They need access to huge amounts of energy,
32:30huge amounts of water. It needs to be a safe, stable place. And then finally,
32:35it needs to be a country that is not subject to the restrictions that US government is putting
32:41on the sale of chips. These incredibly powerful chips are the things that enable data centres
32:48to do their extraordinary work.
32:54In April, one of the world's biggest AI moguls came to town.
33:00Anthropic staged an invite-only forum at Parliament House. The big drawcard was CEO Dario Amadei.
33:09The benefits are real. And if we can mitigate the risks, we can have a much better world. I
33:15really believe that. Just hours earlier, Anthropic's CEO signed a wide-ranging memorandum of
33:21understanding with the federal government.
33:26Good morning.
33:27Three weeks later, Microsoft's CEO announced what he called the largest ever company investment
33:34in Australia.
33:35We're making our biggest commitment in Australia of Australian dollars of 25 billion.
33:44You met with both Satya Nadella and Daria Amadei. What do they want from Australia?
33:49Well, they know that Australia is a very good place for data centre investment. We have enormous
33:57competitive advantages that a lot of economies around the world don't have. A future competitive
34:03advantage in clean, green energy. We have a strong internal consumer market ourselves,
34:09but we're adjacent to the fastest growing markets in the world.
34:15And that means more data centres.
34:19So these are the data halls. This is where the compute that runs a modern economy lives.
34:27Belinda Dennett says Australia is not like Virginia. The market is smaller, the rules are different,
34:33and data centres pay up front for their electricity costs.
34:38Data centres have always been required to pay for electricity connections. 100% of those connection
34:45costs and augmentations to the network up front. That hasn't been happening in the US.
34:55Data centres generate extreme heat and need an immense amount of water to cool them.
35:01And power is another concern. Data centres already consume vast amounts of electricity,
35:08and demand is expected to rise sharply.
35:12I certainly know that we are using more energy, but the data centre industry is committed to meeting
35:19that challenge in the ways that it can.
35:24If big tech has its way, there will be an even greater impact on our energy stocks.
35:31Sources close to the industry have told Four Corners that senior Anthropic representatives
35:36disclosed to them it has ambitions to build data centres that would use a large proportion of
35:42Australia's electricity output. They said Anthropic disclosed it wants five gigawatts of new compute
35:49to help train its AI models by 2030, with a longer-term goal of 20 gigawatts.
35:57Enough at full power to increase Australia's annual electricity generation by around 60%.
36:05Anthropic would not comment on those figures, but said in a statement it
36:10wants to be an accelerator and a partner in Australia's energy transition.
36:14We plan to scale our demand, including through new generation, cover our own costs,
36:21contribute to grid resiliency through load flexibility, and partner with the community
36:27to ensure the benefits are broad and shared.
36:30The federal government said discussions are commercial in confidence.
36:37In Anthropic's MOU with the government, it states Anthropic recognises the importance
36:43of expanding Australia's energy supply and transmission with a focus on firmed renewables.
36:50But the agreement is not binding.
36:54Why didn't you make them guarantee to build renewable to power their data centres?
36:59Well, it sets the direction. This is an early, clear signal
37:04from the Albanese government and the data setter expectations.
37:07The early engagement with potential investors is setting very clear parameters,
37:13very clear expectations for future investment, but there is always more work to do.
37:18Are you familiar with it, Hannah? I am not.
37:20For Helen Toner, the case for data centres is not just economic, it's strategic.
37:27I think building a lot of data centres in Australia could be a great idea for Australia,
37:32if it's done right. I think in the 21st century, computing power is going to be
37:37a major source of national power. And so, if Australia is concerned about being a little bit left
37:43behind in the AI era, then making sure that we have a lot of that computing happening in Australia
37:48will give us leverage potentially in the future because it's such an important resource.
37:57As the Australian government attempts to make the most of the productivity benefits of AI,
38:02it's coming under fire from one of its own over AI safety.
38:08I think we run the risk of terrible judgement potentially from future generations that think
38:14that we had an opportunity to get this right, we had an opportunity to rein in risk
38:19and to really get the best out of AI and we didn't take it.
38:25Morning everyone, thanks for joining us. In 2024, the then Minister for Industry,
38:31Ed Husic, announced the Albanese government had plans to make AI safer. He wanted mandatory guard
38:38rails to better control high-risk AI, including minimising harm to children, protecting physical
38:44and mental wellbeing and preserving civil liberties and judicial rights.
38:49Our proposed policy was that the developers of the technology, the deployers of the technology,
38:55think about risk, prepare their approaches to deal with that risk and to build confidence
39:01in the minds of Australians that they could use AI safely.
39:06Ed Husic had consulted widely with industry and experts, but after the election he was dumped
39:13from the ministry. By the end of the year, the government had abandoned his plans.
39:22Today we'll be launching Australia's National Artificial Intelligence Plan.
39:28The new minister announced there would be no mandatory guard rails on high-risk AI.
39:33The National AI Centre released a series of guard rails,
39:41not mandatory guard rails, but guard rails for business adoption.
39:46We've done a 180 degree turn. We've gone from recognising the risk and recognising that we
39:52should do something to now having a sort of hands-off approach. And I just don't think
39:57with a technology that promises to touch every part of our lives, we can have a hands-off approach to
40:03AI.
40:05Ed Husic believes the government backed down out of fear of a backlash from the US.
40:12We blinked in the face of Donald Trump and decided that we just couldn't get away with
40:17our own approach on this, that we were already attracting heat from him and Elon Musk on social
40:23media laws. Did the government back down on an AI act to appease Donald Trump and the other
40:29big tech CEOs in the US? Well, certainly not. We are absolutely prepared to act, to legislate,
40:36to regulate big tech where it's in the national interest. We've moved with world-leading legislation
40:44to limit the access of under-16s to social media. But not AI. We've moved to outlaw
40:52deepfake porn, often enabled by artificial intelligence. We've delivered an artificial
40:59intelligence safety institute. Tell us why we don't need a specific AI act. Well, I'm yet to be
41:06persuaded that an act that's developed in 2026 will be adequate to meet the challenges of 2027,
41:14let alone 2030. The government's national AI plan stresses that existing laws are the best way of
41:23dealing with AI-related risks. What kinds of guardrails and restrictions might we put on AI?
41:33And I hope that we put in the right guardrails for artificial intelligence, especially to protect...
41:38Toby Walsh was on an AI advisory group set up under Ed Husic.
41:42Today, rich people have personal bankers. It was abolished under the new minister.
41:48My fears about the government's approach is that they're listening too closely to vested interests.
41:56This was always true for mining, but it seems to be becoming ever true for the tech industry,
42:02that we see lots of nice announcements where they stand up on stage with CEOs of
42:09tech companies. But I don't see enough critical voices asking about what are the consequences here.
42:15I'd say I welcome a debate in the community about the risks and opportunities. There are risks that
42:21the technology presents. There are also risks about going backwards and about being left behind in
42:29technology terms. And those risks present as economic risks, but they also present as strategic and security risks as well.
42:38Open AI, Anthropic, and Google DeepMind declined interview requests.
42:44Many of those closest to the industry, researchers in AI, those who've worked for the big companies,
42:50or report on them, say action is needed to regulate the risks around frontier AI.
42:57And I think that's the only thing that we've worked for in time.
42:59My true belief on this matter is that this stuff really is dangerous. And that while it does have
43:06many positive applications, I feel like I benefit from AI in my life basically every day now,
43:12it can be really dangerous. And by default, if we don't apply sensible regulations and sort of get some of
43:19these companies under control, I think we could be in a lot of trouble.
43:29I think we have a pretty difficult challenge ahead of us. But what makes me hopeful is we see people
43:35in our political system on both the left and the right taking these concerns very seriously. We see
43:40more and more people talking about this in public. And like, we need good governance for this to go
43:46well. That gives me hope.
43:48And it's not too late?
43:49It is not too late. We are not yet at the point where these AI agents are superhuman at the
43:56things
43:56that matter most. They're not yet superhuman at persuasion. They're not yet superhuman at strategy.
44:01We have an opportunity now to intervene before we get to the point where it's irreversible.
44:07But we might not have much time.
44:43If this program has raised concerns for you, you can contact one of these services.
44:47Don't forget to subscribe to us.
44:48We'll see you next time.
Comments