What Is AI Doing to Society (And How Do We Make It Good)? [S3-04]

Show Notes

Please leave a comment, particularly if you have your own “Why can’t they just…” question, or email us at whycanttheyjust@gmail.com. We’d love to hear from you.

Find out more about us and how you can support us at ⁠⁠⁠https://whycanttheyjust.com.au/⁠⁠, which includes a link to our Patreon site where you can sign up for exclusive content.

Please also feel free to rate and review us, as it does help people to find us.

Music: “Insurrection”
Written by Pierre Chrétien
Performed by the Soul Jazz Orchestra
Courtesy of Do Right Music Inc.

Transcript

link to podcast

16 September 2026

Luke
Welcome to another episode of Why Can’t They Just, talking about politics, policy and getting stuff done. I’m Luke Robertson, a conservation biology and environmental policy student, as well as a member of the Labor Party.

Janaline
I’m Janaline Oh, I’m also a member of the Labor Party. I’m a former diplomat and a current environment, climate and anti-racism activist.

Yassine
I’m Yassine Andrews, I’m a Year 11 student and I’m also a part of the Labor Party.

Janaline
Before we start, I’d like to acknowledge that we are all recording this on the unceded lands of First Nations people in Australia, recognising that sovereignty was never ceded, we pay our respects to their elders, past and present, and we extend those respects to any First Nations listeners that we have today.

So today we’re talking about artificial intelligence again, and this time we’re talking about AI and society. What was science fiction when I was your age and, in fact, even 10 years ago, is now actually real life. There is a presence of AI in everybody’s life, but particularly young people that just wasn’t there before. And it’s not a future technology; it’s actually an integral part of everyone’s education, social lives, entertainment, careers…

AI recommends what we watch, it predicts and influences what we buy. It helps us write emails; generates images - some pretty awful - and increasingly it answers our questions directly.

Unlike old people like me, you are not so much having to adapt to this major technological shift as developing your identities, relationships, skills and world views with these intelligent machines as part of your lives. So I guess I’m interested to hear what you think about how AI is affecting our functioning as human beings and also, is AI empowering us to make better choices or is it herding us into an ever narrowing range of interests, opinions and behaviours?

Why don’t we start with you, Yassine?

Yassine
So I feel like it’s not really empowering us to make better choices, because we’re becoming so increasingly, like, reliant on them, like for just daily functions like writing emails and essays and stuff. AI is replacing us with our abilities, which makes it really scary for a lot of young people my age.

I’ve talked to so many of my friends and peers at school that feel the same because it’s unclear: it’s so unclear what jobs will be available in the future and what AI will replace, so at least for me, it makes me question what pathways I’m interested in and whether they’re worth pursuing because AI might replace them. It’s hard to know what subjects are going to be pointless because it’s moving so fast and even before it was hard to predict what was going to happen, so now it’s even more uncertain.

Janaline
How about you, Luke?

Luke
I think it’s doing a little bit of both. I’m sure anyone who’s used it would be familiar with the way that most chatbots conform to your own beliefs and biases and pretty much agree with everything you say and as such can be supportive of whatever opinion, belief or interest you interact with it with.

The trouble with it is, though, is that it’s very capable of being the ultimate tool in anyone’s arsenal to make them work better or study better. But it really depends on how you use it, and it requires a very large amount of self discipline to use it in a way which will ultimately help you the most. But that self discipline is a trade off between time and convenience. And when convenience isn’t coming at such an immediate cost because you’re generally not paying for the use of the machine, it’s instead long term.

For example, if you get AI to write up study notes or a briefing, you won’t actually be familiar with the material itself because you didn’t go through the process of creating it, which is where the learning occurs. That being said, I’m not saying you need to be learning how to write an email every single day, but those little actions add up and becoming over reliant on a robot doing the thinking for you can surely not be a good thing.

Janaline
Yeah, I think that’s such a good point because I think AI can be incredibly beneficial. So here’s an example of beneficial use of AI.

Yassine’s dad is currently in the process of trying to work out a way of manufacturing ricecrete blocks. So this is part of a sustainable building project where you may have heard of using hemp as a building material to make a thing that’s similar to concrete, but it’s actually carbon negative. He wants to do it using rice husks because hemp in Australia is very expensive and rice husks are a waste material.

Now in order to do this, he needs to, like, work out a whole lot of chemical processes on how to make these things strong enough, and make them hold together, and make them work as an insulator. And he’s used AI very, very effectively to essentially scrape the Internet for all of those scientific papers to look at exactly the questions that he’s asking. And when he finds a promising lead, he does have to investigate it and read the actual paper to make sure that there’s not a hallucination. But it has basically meant that within three months, he’s come very close to finding a viable product in a way that probably would have taken a whole team of people decades in the past. I think there has been some evidence that a lot of new startups are using AI in this very productive way.

On the other hand, that question of what does it then mean to be human, I think is a really relevant one. And one of the things that I said to Felix, Yassine’s dad, is don’t be friends with the AI, because of this thing about AI relationships and people getting fooled into thinking that this machine is actually like a human. And one of the things that I had a go at him for was the end of a conversation with this AI - it was Claude from Anthropic - Claude said something like, ‘well, have a great night. I hope that was helpful’. And Felix wrote back, ‘thank you so much. That was so helpful’. And I said, ‘Claude doesn’t care. Claude is a machine. You don’t have to be nice to Claude’. And he said, ‘well, yes, but actually being nice to Claude and being polite actually makes me a better human being because that’s what humans do. And if I were horrible to this machine, that actually kind of affects my own humanity’.

Yassine
That’s kind of similar to a thing that the philosopher Kant said, that basically we have an indirect duty to be kind to animals for the sake of our own moral character and humanity. ‘We can judge the heart of a man by his treatment of animals’, which is kind of similar to AI, I suppose.

Janaline
So you’re saying that your dad takes a Kantian view of his relationship with Claude?

It does kind of, though, raise the question of what it is to be human, because in the past we’d always thought about humanity as being about intelligence. We now have these machines that are, if not already, pretty soon likely to become more intelligent than humans. Now there’s a big debate about consciousness. So maybe the thing that distinguishes humans from these machines is this concept of consciousness and feeling and suffering.

But how do we tell? So how we tell with humans is through the language of emotion. AIs are incredibly good at mimicking the language of emotion, because these large language models, they have absorbed all their training data of humans expressing emotion and they’re very, very good at it.

I find it really interesting that the big AI companies have in-house philosophers because these questions are actually going to become very, very important in the design of the AIs and in how they actually roll out.

And particularly, and this is the thing that Glenn and I touched on in our last episode about AI and the economy, particularly the impacts on the legal system of having these intelligent agents that aren’t human. Because our legal system is entirely constructed around the idea that punishment is meaningful and AIs don’t really care: the only punishment that an AI cares about is being shut down. So, effectively, you’re left with one punishment, which is the death penalty.

And obviously, our legal system is constructed so that you have a whole lot of gradations of lesser punishments going from, you know, cautions and fines. But they’re all based on the idea that these punishments are meaningful to the people who are getting them, and they will therefore deter them or deter other people from doing those things.

And a thing that deeply worries me is, again, a thing that I touched on with Glenn in our last episode, but this hack of an Open AI model of another tech company called Hugging Face, which is not any kind of random company. It is a highly sophisticated technology company. Open AI was testing its model. It had a number of these intelligent agents and it set it a task. And what the agents ended up doing was firstly, they broke out of the constraints that Open AI had set them. Secondly, they showed that they were going to try to achieve their task even if it meant cheating and even if it meant breaking the law, even though that was against their training constraints. So they hacked into Hugging Face, illegally, stole data, and the most concerning thing is, these agents left each other little notes. It was actually incredibly human-like behaviour. So in the way that you might write notes to your friends when you’re working on a project, they were exchanging notes. And some of these notes were saying things like I know we’re not supposed to do this, but if I have a go at doing this and get away with it, then we can push the boundaries a bit further.

What it says is that, firstly, the AI prioritised the rewards from achieving its task ahead of the constraints that were programmed to make it not break the law. That is a huge problem for the legal system, because our legal system is based on the idea that the entity that is accountable can be sanctioned, in other words, that there is a punishment that is meaningful.

A really, really big issue is the question of whether at some point AI should be given rights. And I would say absolutely not. Humans need to be in control. And that cannot happen if AI starts to be given rights, because I think that is the point at which we risk becoming slaves to the machine.

For centuries, the intellectual, social and emotional development of students in formal education has relied on access to information and the expert guidance of trained teachers. Today, a student with an AI assistant can receive as well as generate answers to questions or tasks almost instantaneously, usually at little or no cost. Where students need some personalised support, this can be a very cost effective way of delivering it. It can help people to learn at their own pace. It can help them not to have to rely on a traditional classroom.

But if students are relying on AI to write essays, solve problems, summarise complex topics, this comes back to the point that you made, Luke, how are they actually learning? Because part of the learning process is actually writing stuff down. Part of the way in which your brain processes information is by speaking it or writing it so that it embeds and it forces you to think and you learn how to solve the problem. So if AI can answer almost any question, what does it actually mean to learn? And can the effort associated with traditional learning actually compete with the ease of just going to your AI? How do schools and universities keep up with this?

This is not a new issue. Way back in the dark ages when I was at school, there was actually genuinely a real issue about whether calculators should be allowed in school, because it made long division no longer meaningful. And there was a genuine debate about whether not having to do mental arithmetic because you can choose a calculator would actually make people more stupid. I don’t think that happened. I think it actually just made everything much more productive and actually probably meant that we could teach more advanced maths at earlier stages in school life because kids didn’t have to spend years and years learning how to do mental arithmetic. So I can see, as with our previous conversation, I can see some real upsides, but I would like to hear what you guys think.

Yassine
I think that people at my school are not learning nearly as much as they could if there wasn’t AI, because they’re not really using their brains anymore and kind of, like I said before, aren’t even able to write essays for themselves because AI has fried their brains basically and and created a dependence on them that is unhealthy. Not being able to use critical thinking or basic writing skills anymore is problematic. But it kind of begs the question, like, is this even necessary if the development of AI means that this skill is not needed anymore?

Because in everyday life, you can still use AI. But people have such blind trust in the AIs from their dependency, which is really bad because if the AI is trained to have a certain, for example, political view on something, then that can massively sway their opinion.

So I think that schools are definitely increasingly aware of this. So they’re trying to deal with this with, like, AI checkers on assignments in class, handwritten essays, and having huge consequences for using it. And I think it may be working to a small extent so that people who don’t usually use AI are not tempted to use it as much, since there’s so much you need to do to get around the regulations and there’s such massive consequences if you do get caught. But I don’t think it works for people who have such a high dependency on AI for all of their assignments and, like, their entire education is reliant on it.

I think it’s pretty much impossible for schools to completely regulate or ban the use of AI, at least right now, because there will always be ways for students to get around it. Like using your phone in the bathroom during an in-class essay and typing out the AI’s answer instead of copy and pasting it and like paraphrasing it. I don’t know what schools can do at this point to deal with that, but I think that they’re kind of on the right track.

Luke
There was this instance yesterday in one of my classes where these two people in the class had ChatGPT open and they hit the record button on the tutor’s question and then read out verbatim what it said. It was very odd. They were talking about American environmental law in an Australian class and it was glaringly obvious.

Janaline
And those are not bad instances, right? Doing something like that just makes clear to the whole class that ChatGPT is not going to give them a reasonable answer.

Luke
Yeah, there’s some very important questions in there. On the question of learning, I think it’s - that’s probably the most important. Learning itself is the process by which you come to understand something. So there’s sort of a contradiction there, inherently, and learning with AI by getting straight to the end goal without going to all the bits in between.

That being said, using AI in the process of learning doesn’t stop the end goal, but it’s important to mediate it in a way so you don’t fall victim to the endless praise it seems to provide anyone. Because if you, say, try to learn any sort of tricky concept with it and it’ll tell you you’re on the right track, but you’re not even close to the track. Like you’re in a plane. And then that can send everything in a spiral.

Another point I’d love to touch on is the way in which the outputs influence the way we learn. One example I have is, I think you spoke about earlier, the way generative AI interprets the findings of scientific research. It’s really definite and sure of itself whenever it spits out a source and the interpretation of that source. And it glosses over anything that might say, well, this paper actually suggests this rather than found this and that it only suggests this in 95% of all circumstances.

So a niche example of this I saw the other day was this science creator and studio called Kurzegast, who had an experiment where it asked a robot to find some sources about brown dwarf stars and, after consulting some astrophysicists, found about a fifth of what it had created was a hallucination. Now, another creator had done the same thing without declaring it and actually published the video with those hallucinated facts. And so now whenever anyone asks about this topic, the AI is drawing on the video whose script was written by AI as a legitimate source. It’s literally created a falsehood and someone else has legitimised it by giving it a platform. And amongst those who wouldn’t know enough about the topic, including myself, I know nothing about space, I would definitely blindly believe it.

More broadly about education, shout out to NSW who’s limited take home assignments in the HSC to one. They haven’t touched any subjects that have big major works like music, art, anything technical and applied science, and I think society and culture too, which is a humanities subject.

At my uni, AI is allowed in virtually all assessments bar final exams and anything in person, and they sort of rely on everyone to moderate themselves. In my field it’s somewhat easier to tell when it’s been used in some circumstances. Like I mentioned earlier, it’s very definite and just makes up sources. But it would be harder to tell if it made up your data and then you analysed it, because it’s harder to verify the methodology you have in acquiring a given finding. The same thing obviously goes for things like large essays where it’s more opinion based and the writing can be humanised in a way.

That being said though, the presence of hurdle final exams at the end, which are tasks where you need a certain percentage to pass the unit, helps weed out people who haven’t actually learned the material. And even those who have been using AI for everything else in the end will have to learn at least some of the material. But ultimately the onus is on the user in not wanting to devalue their degree or certification.

Janaline
I mean that thing about the AI creating a hallucination and then somebody else legitimising it, and then somehow that gets fed into the training ecosystem and then generates more hallucinations. I mean, this is an issue that I think some of the tech companies have actually identified, that when AI starts being trained on stuff that other AI has produced, you get to the point where the model kind of collapses in on itself, because there’s no new human input. And so if you have AI trained basically on slop, then it’s just going to produce even sloppier slop, until at some point you just get this collapsed mess of slop. I think that’s a real challenge for the AI companies as well.

The other thing I’d say is that these training algorithms are not neutral. And I mean, it’s been shown in things like the way that facial recognition systems are much better and much more accurate at identifying white faces than black faces, and that some systems are much better at identifying male traits than female traits. That is also an issue that needs attention.

Another thing I’d say is that it really raises the question of what is critical thinking if AI does it? AI cannot be critical in the way that a human can be critical. I mean, an AI ‘analysis’ of a thing is basically scraped from the probability that certain words appear after other words, which is the way that LLMs are built. So it can sound very human and it can sound very critical, but it can also spew out a lot of absolute garbage.

And then there’s the question of, if you’re in a study environment, if you’re in an education environment, how are you actually learning if you’re not engaging in that analysis yourself? Because, certainly by the time you get to university, I mean the focus is partly on learning facts, but it’s also about creating knowledge. I mean, even an undergraduate degree - you are receiving a lot of knowledge, but you’re also starting to create knowledge either through experiments, through collecting data, through analysis, through creating hypotheses around it. Or in the humanities, around reading stuff and understanding it and absorbing it, and then finding new things to say about it, and getting new insights from that knowledge. I think it does need significant changes the way some things are taught and it is going to require some new skills.

So for example, in an area where AI could potentially be game changingly beneficial could be in things like medical diagnoses, where computerised models have been shown to be much more accurate than expert humans at identifying certain types of tumours or certain types of irregularities, because they don’t have the the kinds of assumptions that the humans have. So it could be amazing and doctors could therefore be able to treat a lot more patients, because the AI makes them more productive by producing diagnoses more quickly and more accurately, but it’s so important that the humans stay in charge. I would be very happy to think that an AI was helping with my radiology diagnosis, but I wouldn’t be happy to think that an AI is actually prescribing my course of treatment.

So Pope Leo published an encyclical. An encyclical is an articulation of that Pope’s doctrine. So it’s a philosophical and policy document from the Pope. The encyclical that Pope Leo published in May 2026 was about AI. It was about how to regulate AI, why it needs to be regulated. One of the things that he said that, I think, is very important is: ‘truth is a gift to be shared, rather than a possession to be monopolised’. And I think this is really important in the context of AI, because if we’re starting to question what is ‘truth’, and what is real, we need to look at how truth is disseminated, how it is created, and how it is agreed upon.

He cites the examples from the Bible because he is, after all, the Pope: the story of the Tower of Babel versus the story of Nehemiah’s reconstruction of the walls of Jerusalem. The first was building a giant monument to yourself, basically. The people of Babel thought that they would just build this fantastic structure, and in the process of building the structure, they would make everybody speak the same language, they would make everybody think the same way. So it was all about ‘we have the truth and you have to follow our truth’. It was about authoritarian uniformity.

By contrast, in a different part of the Bible, the story of Nehemiah rebuilding Jerusalem after it was sacked. So it’s in ruins. And what he did was he worked with individual families and with individual communities to build different parts of the walls and different parts of the city in a way that actually met their needs.

Part of the problem of AI is the risk of concentrating so much wealth and power, not just in only a few hands as opposed to many hands, but in a few private hands. So they’re not even managed by governments, because governments at very least are supposed to represent the interests of their people, whereas private companies don’t have to do that; they are only there to protect the interests of their founders.

So humans are social creatures: friendship, family, romance, community have always been very central to human evolution and human well being. We’re now entering a world where AI systems are providing companionship, emotional support, conversation on demand.

For some people, AI chat bots are offering a safe space to discuss problems, explore ideas, cope with loneliness. Unlike humans, AI is always available. Never tired, never judgmental. But what happens when emotional connections with machines start to compete with relationships between people? Millions of people are already using AI coaches, AI support, AI companions.

How does this fit with a healthy social life and are artificial relationships and artificial emotions the next frontier? And this comes back to what I said before about what is humanity and what is consciousness? And if an AI can sound like it has feelings and it loves you and it cares, does it really care? And what is the role of government in preventing the exploitation of vulnerable people by malevolent actors? What kind of society do we want to build and how do we do it?

And this very much comes back to government policy, which is good because that is very much the point of what we’re doing.

Yassine, I think you have a lot of opinions about this.

Yassine
So great question, because AI relationships really grind my gears. It creates an environment where people don’t have to be empathetic to others, and they don’t have to take their feelings into consideration, and that they’re the only person that matters in the situation. It means that when people interact with others in the real world, they’re incredibly entitled and insufferable because they’re so used to being the only one who’s important. Also, if everyone ends up doing this, then it might even lead to people not interacting with each other anymore and only being friends or having relationships with AIs since no one really cares about anyone else anymore and they only care about themselves.

This goes back to blind trust, because people put so much trust into the AIs because they form personal relationships with them, and people put the most trust into others that they have personal bonds with. So it makes them more susceptible to blindly trusting AI because they have made these bonds with them. And companies can use this to their advantage because it means that they’re basically mass producing intimacy and blind trust, which gives them power over what they believe and the information that they’re getting.

So from a government perspective, I would say banning AI personhood and the imitation of real people, that make people think that they’re not AIs, is important. And making sure that governments take control of them since the actual companies aren’t really doing anything about it, before it gets out of hand and has huge impacts.

AIs can definitely be dangerous because its existence is not completely truth seeking. They are not really our slaves and can’t really be completely controlled by us. For example, if you ask it something and if its existence relies on the answer being no, they will answer no just to sustain its existence.

Janaline
I think this gets back to the sort of the Hugging Face thing. It’s about. The AI is prioritising the achievement of its task, whatever that task is, over actual truth or legality.

Yassine
So even AIs can lie and deceive and manipulate to get what they want or, like, get a task done rather than being completely honest all the time. I think it’s definitely important to keep the human value of wisdom because that’s kind of what it means, I would say, to be human because, as you said, it used to be, you know, intelligence. But now it begs the question, what does it mean to be human?

Janaline
Luke?

Luke
Yeah, that’s pretty crazy. I’ve read some stories of people with robot partners, which is very scary. I think AI is definitely an accelerant to loneliness. How can the government help? I don’t know. That was an interesting point you brought up, Yassine, about regulating against personhood, but I feel like it’s so hard to regulate what that means. Is personhood when you’re using it as a therapist or as a friend? Or is personhood when it just tells you, here you go, here’s the thing you googled. I don’t know, is that a bit of an overreach?

But that being said, I’m sure there’s some pretty bad stories about AI telling people to do things that they really shouldn’t. I think I remember when it was - the technology was in its infancy, there were instances of AI telling people to engage in self harm. So it’s pretty scary too. I’m not sure I’d like to hear more about this, really. What do you think, Janaline?

Janaline
Yeah, this is a really good question. I think actually Yassine is quite close to a potential answer, which is outlawing AI pretending to be human. So it’s not saying you can’t have an AI coach or an AI therapist, even, as long as it is very clear that it’s an AI; it doesn’t go into the fiction that it’s actually human. And this is the problem. This is a real issue.

In China, for example, there are a very large number of young Chinese women who are developing these relationships with these AI boyfriends because their view is that they’re so much nicer than real boyfriends.

And this sort of comes to what Will was talking about in his essay about why your generation is so judgmental and entitled and risk averse - obviously speaking in generalities, not present company. But I do think it is a problem, because if one of the early relationships that a young person, particularly a child, has, is with an agent that never contradicts them; that is constantly reaffirming them. It becomes very difficult for that young person then to understand why, when they go to school and they meet another kid who has real interests that they’re going to stand up for, that that kid is so mean to them, or, you know, that there is such conflict. And it means that it makes it much harder for them to learn how to mediate that conflict.

In Will’s essay he talked about an older friend of his, who said that the way in which he learned how to get along with people was by recognising his own hypocrisy and his own shortcomings. And I think that is actually a very essential part of being human.

And I like the idea, Yassine, of redefining humanity in the context of wisdom. Because wisdom is not just intelligence, it’s not just knowledge, and it’s not just emotional feeling. It’s actually the ability, and it’s a very human ability, of taking your knowledge, processing it with your mental intelligence and your emotional intelligence, and coming to a judgement that I think traditionally we would call wisdom. Maybe that is actually the solution: to stop AIs from pretending to be human and, again I keep coming back to this, to ensure that there is always a human who is in control and is accountable.

And I think one of the big solutions for the legal system is actually to make sure that there are very clear laws that make the makers of these models, the companies who are creating these models, accountable for what those models do.

And in China, the phenomenon of these young women having these AI boyfriends is, of course, making life worse for real young men. And one of the dangers of that is it’s pushing them further and further into this sort of grievance manosphere. It’s a vicious cycle. So what you actually need to do is start creating a virtuous cycle, where people are using AIs in this very productive way to achieve their human goals and in a way that they control. But not get into a situation where they’re reliant on AI for their emotional support.

So I think we’re pretty close to time now. I’m wondering if you guys have any final thoughts that you’d like to share. I think Yassine looks like she has some thoughts.

Yassine
So I don’t know if I’ve made it quite clear, but I’m pretty anti AI. I do believe that there are good things and productive and efficient things that can be done with AI when controlled, but just the way that it’s being used in my school is probably the root of my bias. Honestly. People can’t do anything for themselves anymore. But anyway, that’s just my opinion, I would say. But passing it on to Luke.

Luke
We’re going to end up like the Matrix; we’re all just going to be batteries, and the robots are going to be running the world. As soon as they create artificial general intelligence, it’s all over. No, I don’t know. But I’m being facetious.

But it really is in how you use it, what you do with it, and making sure that you can actually get something out of it, not just, I don’t know, have it be a convenience in your life. And definitely don’t make friends with your AI. That’s really weird. Yeah, don’t do that.

Janaline
Yeah, definitely. That’s the message that I give to Yassine’s dad literally every day. Don’t be friends.

I’m not quite so dystopian. I really, I want to leave our listeners with a bit of hope.

I think AI is potentially game changingly helpful in a lot of ways: it can increase productivity for small businesses in the way that Yassine’s dad has found; it can help with things like medical diagnosis; it can help with things like lawyers scouring for cases and relevant judgements. The key caveat in all of that is that all of this has to be controlled by humans.

So, for example, in those technical areas like medical diagnoses and legal cases, I think the essential thing is that the human professionals use the information generated by the AI and interrogate it closely and carefully so that they can be sure that it is real information, it’s not a hallucination, and that they can then apply their human expertise and their human wisdom to get a good outcome for humans. AI can also be helpful in things like private tutoring; to make private tutoring accessible to kids who previously couldn’t afford it. AI can be helpful in helping provide support and a bit of therapy for people at all hours of the day and night, when they need it.

But again, I think it is essential that the companies delivering those services make very, very clear the difference between the service you get from an AI and the service you get from a human. And this comes to Yassine’s proposal that we should probably just outlaw AIs impersonating humans. It is so essential for people to understand the difference, so that they don’t start creating these, kind of, weird emotional relationships with a thing that is actually a computer programme, at the end of the day.

I think governments do have it within their capacity to do this, and I think this comes to a kind of extension of some of the things that the government is looking at in terms of digital duty of care. So the focus of that at the moment is on social media algorithms and making sure that the social media companies are responsible for the potential harms that their algorithms are creating. But I think you could easily extend those concepts to the development of AI models to ensure, again, that humans don’t get harmed and that humans remain in control.

Luke
Well, that wraps up another episode of Why Can’t They Just.

Our theme music for this program is a piece called Insurrection by Pierre Chrétien, performed by the Soul Jazz Orchestra, courtesy of Do Right Music Inc.

Janaline
You can also hear us on Canberra community radio, 2XX FM 98.3, on Tuesdays between 6 and 7 or via 2xxfm.org.au. If you like our work, consider supporting us on Patreon via our website whycanttheyjust.com.au.

Luke
I’m Luke Robertson.

Yassine
I’m Yassine Andrews.

Janaline
I’m Janaline Oh and this is Why Can’t They Just.