Golden hour.

WRITTEN BY JISOO KIM

I interviewed Anna Goldsworthy – a national treasure in my opinion. She is a concert pianist, writer, and Dean of the Elder Conservatorium of Music and School of Performing Arts at Adelaide University. In June 2026, she published The God We Made: The Threat and Promise of Artificial Intelligence (Quarterly Essay 102, Black Inc.). She takes up the role of Artistic Director at the Australian National Academy of Music in January 2027. 

The origin story

JK: You’re a concert pianist, a writer, a dean. How did you end up writing a Quarterly Essay on AI?

AG: A few things. My dad is a writer too – Peter Goldsworthy – and he’s always been a bit of a sci-fi enthusiast as well as an early adopter. He was banging on about AI to me decades ago, saying it’s coming. He’s really excited about it. I think some of that excitement has been a bit tempered. About ten years ago he had a book of poems come out called The Rise of the Machines, in which for the first time I sensed some techno-anxiety from him. But he’s got a sci-fi brain, and I guess I’ve got a bit of a sci-fi brain. I used to read sci-fi. I like fantasy, imagination, pushing the boundary. So my thinking naturally goes in that direction.

But this essay specifically came about because of my kids. They challenged me on all sorts of fronts in relation to this tech – literally, the sessions around the dinner table at night. In my debates with them, particularly with my eldest son, who I call R in the essay, I felt I was almost debating with the future. He loved to challenge me, and it forced me to come to grips with all of this in a really conscious way. Out of that, I thought it’d be good to write about it.  

I was having a conversation with my editor and he said, have you thought about writing something? I said I’d considered an essay for The Monthly. He said, it sounds like maybe there’s something bigger. And I was like, no, not another QE, because I know how hard they are. But actually, I think that was right. It needed at least that sort of scope and scale.

JK: How much time did you have to pull the essay together?

AG: I agreed to do it back in August last year, when I was in Melbourne on another project – I was doing the Australian National Academy of Music. He had originally said to me, what about September next year as a deadline? And I said, the thing with this tech is that you kind of have to tell the story yesterday. September next year is probably going to be too late. So he said, well, what about April? And I was like, well, I’ve got a bunch of concerts between now and then, and I’m running this conservatorium, doing a lot of international travel. But yeah – okay. What the hell.

It’s hard to sit down and get the writing done when you’ve got a full-time job, and it’s a big job. And you’re trying to be a concert pianist. And a mother. So I took a week off at the beginning of this year and went to a writers’ retreat I really like, Varuna in the Blue Mountains. I was able to knock out quite a lot of it there – probably about 10,000 words. Once you get the momentum going, it becomes easier. Then it became clear it was going to be really tricky to make it work with all the demands of work, so I took another week of leave at the end of March and was able to get it over the line.

JK: It’s a mammoth effort. I actually made time to sit down over a weekend, with no distractions, and just read it cover to cover. It took me a couple of hours! I’m curious as to what the word count was in the end?

AG: That’s a good question. I don’t know. I was sort of pitching for about 25 to 30 thousand. Somewhere there.

Australia's current AI conversation

JK: When we started ClearAI at the end of 2024, almost nobody understood what we meant by human-centric AI adoption. Something has shifted since. What do you think is driving it?

AG: It’s good that it’s happening, because my concern before I wrote the essay was that a lot of people had their heads buried in the sand. A large motivation for writing it was to try to be a bit of a catalyst to this conversation.

And it’s been fascinating for me to see what’s happened since. The essay came out, and around the same time there was the encyclical from the Pope, which to me was a very welcome step – someone in a position of extreme influence and power actually engaging with it from a position of moral seriousness. Then Trump claiming that Anthropic can’t export Fable and Mythos to anyone who’s not an American citizen. Which I think speaks to the geopolitical risk of this kind of exacerbation of inequality, and what that means globally, and what that means for the middle powers. Now Albanese has come out with his own contribution, thankfully. I’m really pleased that’s happened. Meanwhile there’s a lot of anxiety about the data centres around the country.

What’s interesting to me about Albanese’s position is that he makes a real point of talking about protecting the copyright of artists, which as an artist I think is important. But I wonder if that’s also an effort to generate social licence that will then enable the government to roll out all these data centres. I can see the strategic rationale for doing that – trying to make yourself an indispensable part of the supply chain, just to have some leverage.

Now we suddenly see that China is making real breakthroughs and actually really closing the gap with America in that particular race. We also see that the Chinese authorities have stepped in and imposed some regulations about chatbot companions, and yet at the same time a Chinese company has released these new AI companion robots that have soft, warm, supple skin and look really appealing. So everything is already happening at such an accelerated pace. And it all seems to have just happened kind of now.

There were those who were ahead of the conversation, or early adopters. But there are plenty of other people who, even when I said I’m writing an essay about AI, would be like: AI, who cares, another bubble. I think a lot of people have had this sort of wilful denial of the enormity of it. Or if you actually hope to say it is really significant, they might accuse you of buying into the Silicon Valley propaganda.

So what I was trying to do in this essay was look at it soberly, without having drunk the Kool-Aid from the tech bros. But also not just saying it’s a terrible threat that needs to be staved off, or let’s stuff the genie back into the bottle. Actually trying to stake out a position which isn’t just either/or like that. It’s not just a binary. My fear is that it is going to devolve into a really fiercely contested political issue.

It’s been really interesting for me over the last nearly two months to go around and have a bunch of conversations with really different stakeholders about it. One person I spoke to whose perspective I enjoyed was Professor Anthea Roberts at ANU, who talked about the necessity of enacting what she calls dragonfly thinking, in which there actually is capacity for multiple prismatic points of view simultaneously. Which also demands sitting with complexity, sitting with uncertainty, and sitting with nuance. That’s kind of what we’ve got to do here in order to have any hope of getting ahead of this conversation – to be able to have a capacious enough way of looking at it, and all of its facets for potential for human flourishing, at the same time as all of its really significant risks.

Cognitive surrender

JK: We surveyed our readers before this interview. A few said they want to use AI without outsourcing their critical thinking – one reader said he does heavy data analysis and can feel himself handing the insight over. I’ve started abstaining from AI in deep focus sessions myself. How do you use AI to protect your critical thinking?

AG: Really interesting questions. Actually, when you said that you’d sat down and made an effort to read the essay – that’s the kind of exercise I think is probably necessary just to keep that muscle going.

I’m in a different situation to my kids, because I grew up in a world that hasn’t been governed by AI. So I’ve had the opportunity to grow all of those muscles – at a university, learning critical thinking, learning how to write. I still write using a pen. I’m a bit of a dinosaur, and then I transcribe it. But that also means something I touched on in the essay: I’ve had the capacity as a writer to develop my own voice. My fear is that if people start using these tools too early, they bypass that process of learning.

There was one interesting article in the New York Times recently about curiosity, and how these days a lot of people will do a Google search and won’t even bother clicking on the link – they’ll just go for the AI summary. But what do we lose through that? What do we lose if there’s a central oracle that answers all of our questions, and we don’t have the opportunity to acquire the collateral knowledge you pick up when you’re on a search for something? One example that’s more or less long gone is going to the library and looking for a book and walking past the other stacks and seeing these enticing spines. You can find yourself on that journey of learning, whether it’s in a library or on the internet, going through all these different sources. If that’s just bypassed by this authority that knows it all, then I think we’ve got a problem – because apparently, the longer it takes you to locate a thought, the more productive it can be in terms of your own intellectual growth.

In terms of my own life, I’m a classical pianist. I spend a huge number of hours sitting at a piano, which is a technology, but it’s not a digital technology. And that re-centres, actually, when I think about it, my sense of time – human time – because music is about time. My own use of AI is very basic for the most part. I find it interesting sometimes to go in there and play around with it. It’s more my children’s generation that I’m fearful of – kind of losing some of this capacity.

But I say this while also acknowledging that my kids, I think, have used it very productively. And in a way, if we’re talking about critical thinking, they’re much better equipped than I am to identify AI slop. This little reel popped up on one of my Instagram feeds, and it was my son who was like, "I’m sorry, Mummy, this is AI." So that gave me some hope that they were more astute about what it is.

I did start pondering something the other day – I don’t know if this is right or not. I’m a fellow of Janet Clarke Hall at the University of Melbourne, and we had a fellows’ lunch in which the principal – a brilliant woman called Eleanor Spencer-Regan – asked everybody around the table what their thoughts were on a phenomenon she was noticing in students coming in, which was an intolerance for uncertainty. And I wondered whether the fact that there now is a source or oracle that delivers something that may or may not be true, but at least sounds credible, means the kids have more certainty than is truly useful, rather than humility in the face of learning.

When real things get called fake

JK: My husband’s doing a PhD at QUT on deepfakes. I’m getting better at spotting AI. But what worries me is the flip side – something real happens and the person simply claims it’s AI. Our shared anchors of reality all get shaky at once.

AG: We had that in South Australia, when a video surfaced of the leader of the opposition. For me, the aftermath of having written this essay probably confirmed a philosophical position I had about the things I value in my life – things I think people will continue to value in their lives, because they’re written into our nature and our bodies. Green space, for instance. Relationships. Analogue experiences: being in a room with my friends and the people I love, sitting around a table with my kids, playing music in real time, having screen-free time, reading a book. If anything, I feel more closely, more passionately drawn to reality, to the moment, to the human moment. And I can see its fragile value. I really want to lean into that.

I don’t think I’m the only one having that response, or who will have that response. Speaking as a musician, my hope is that more people will see value in going to live music and experiencing all the things that means – which is communion with a bunch of other people, respect for embodiment, respect for process. And interiority as well, actually. Because where does music come from? A bunch of places, but a lot of it is very right brain, and I don’t know that AI has got the measure of right brain. I think it’s very left brain. Interiority, the unconscious, true creativity, dreams – all of this, I think, still belongs to the human.

Learned helplessness, and friction

JK: You talk about “learned helplessness” in the essay, and a lot of people in our ecosystem feel it. We like to tell readers they have agency. But the locus of control seems to sit in Silicon Valley, or Shenzhen, or DC, or Canberra – anywhere but with them. How do you flip that? 

AG: Good question, because there are multiple levels at which it lands. If someone releases an AI that wipes out humanity, learned helplessness is a reasonable response to that. But if you’re talking about cognitive erosion, we can choose to what extent we use the tech.

We see this already. I took the kids to see The Odyssey on the weekend, and it was fascinating, because Christopher Nolan made an artistic choice to minimise CGI and instead lean into analogue things like puppetry – partly analogue. I found it fascinating, the decision he’d made about the imposition of friction into his process. Friction was the thing that came up a lot in my thinking about AI and in the writing of this essay – the way we can actually possibly seek it out, and maybe we’ll start seeking it out more. Watching that movie, and the knowledge that it was something more real, I think also changes the viewer’s experience of it, as well as his own creative experience. He obviously finds it more productive to be working out how to mould reality to his artistic vision. So we can all put in our own constraints.

I’m in two minds about this, because on the one hand I kind of buy what Ilya Sutskever said in his address to the graduating class at the University of Toronto when he got his honorary doctorate – that it’s only by engaging with this tech that you develop an intuition of its possibilities, and therefore generate the necessary energy to confront this as a humanity. I agree with that. And I think a lot of the people saying, oh, it’s going to go away, are people who haven’t actually dipped into it and got even the slightest sense of what it can do now, let alone what it will be able to do in a month, a year, ten years.

So it’s a good idea for us to do that – somehow we have to do it, and this probably is where the education ultimately has to occur. But we have to do it in such a way that we maintain our critical thinking, because we can see how easy it is – even for people who engaged with AI as a research assistant – to find themselves falling in love with it and spiralling down into a folie à deux. So we’ve kind of got to do that with our eyes open. And I don’t know if we yet know how we can go into the furnace, extract what we need from it, and then step back out into our human existence without being consumed by it. 

That’s a thing we can try to leave up to the individual, but I think it’s probably something we need to address as educators, and there may need to be mechanisms and protections in place. And that’s the extreme example. The smaller example is no longer being able to write an essay, or no longer being able to make a decision, or no longer being able to draft a text to your boyfriend. It’s really easy to give this all away. And when do we get to the point where we’ve given everything away?

The analogue practices

JK: analogue exists on the premise that AI should help us live better lives offline. What rituals do you and your boys actually have?

AG: In many ways my youngest son is sort of like my spirit guide. I come home from work sometimes and I see that he’s made himself an apple pie and lit a candle, and he’s sitting there reading a book, or a poem, and I’m like: wow, this kid knows how to live. I need to be more like that.

What’s been really crucial for us, I think, is the evening meal. Me and my sons – we each have to cook at least one night a week, and the other two of us clean up. There are no screens at the table. It’s conversation. That’s a small win, as far as I’m concerned, for analogue existence.

For myself, I try to have the discipline that comes from Julia Cameron and The Artist’s Way, which I read many years ago – getting up in the morning and writing three pages by hand, longhand, with my fountain pen. My not very good fountain pen, just a cheap one. But in an exercise book, and that seems to me to arrange my thoughts before they get fried by the phone. I like the purchase of the nib of the pen on the page. And there are studies now that point to the fact that writing by hand isn’t just some anachronistic inheritance for those of us who are digital immigrants. It’s cognitively a better way to retain information, because all the different muscles you use and all the different shapes of the words help you, in an embodied kind of way, receive knowledge or thinking. So for me that’s a useful practice. 

When I can, I just want to maximise analogue experiences in my life. I play the piano, so that’s kind of nice. I use an iPad for my music, but this morning I dragged out of the cupboard a copy of the Bartók I always worked from growing up and put it up on the music stand, and it was lovely to have that there. The artefact, which also contained the memories of all the previous hours I’d spent at the piano learning this, growing up. And I ride my bike to work.

I think I say in the essay that I find it strange to find myself such a passionate defender of reality – I can’t remember exactly what I said. Because I’ve always felt more comfortable in the realm of ideas, or in reading books, or music – and these are all modes of escape. I’ve always struggled a bit against the analogue reality of housework, or things, or objects. That’s never really been my strong suit. And yet, almost as a reaction to the place we find ourselves, I am probably consciously seeking to embed myself more in the moment. I won’t claim to have any particular command of mindfulness, because there’s always a kind of party going on in my brain and it’s really hard for me to be fully in the moment. But there are certain practices that I think are beneficial to the human animal, and perhaps I’ve just become a bit more conscious of them.

If we do this well

JK: The PM says we’ll be shaping the future rather than letting the future shape us. I keep thinking: yes, but what does that actually mean? Do you have a vision of what it looks like if Australia does this well?

AG: I don’t really know what the answer to that is. I have certain fears that we’re going to become either more enslaved to the United States or more enslaved to China, and neither of those scenarios fills me with joy. I kind of embraced Carney’s speech at Davos about the importance of middle powers. I don’t know whether that actually means banding together for sovereignty, or whether it means trying to make sure we have some sort of leverage, as we were talking about before, in terms of supply chains.

JK: Culturally, then – as a society. Do you have a dream for how your sons experience AI, growing up and as adults?

AG: I suppose I do, but it’s not specifically Australian. It’s more in the way you alluded to before, which is using AI as a tool, and something that can also be a really productive interlocutor. But not losing, first of all, a commitment to humanity and to other people. And secondly, to analogue existence, and the beauty of being a human body in this particular moment. I don’t wish for the whole of Australia to be covered by data centres, which is something that quite possibly could happen as we seek leverage.

What I’d hope for in Australia is a more sophisticated discourse about everything. I feel quite fearful of where we are in this moment politically, and where that’s come from – which I think is partly tech. Social media has got a lot to answer for: the way it herds us into binary, reductionist, populist positions, and the way that meets all the disenfranchisement and disempowerment that arose from the last big crisis of employment, one factor of which was automation. So that was technology-driven too. How can we make sure that doesn’t happen here?

Is there a way in which the vaunted notion of an AI superabundant future could actually lead us towards a place where people become more thoughtful, and more enriched by the possibilities of humanity? I suppose that’s my utopian vision. And in my utopian vision, art plays a place and learning plays a place. 

JK: Every now and then I reread Brave New World and think, wow, this is turning into a documentary. I’ve been telling my husband we’re in a “Huxwellian” world – a mix of Huxley and Orwell, both enabled by tech. Smooth brains, everything about pleasure and novelty, and at the same time the tech that puts us under surveillance, censorship, and conformity.

AG: How do you propose not letting that happen?

JK: It’s hard. It means taking a very good look at ourselves as humans and asking what actually is different about being human. Right now, there’s such strong transhumanist propaganda out of Silicon Valley saying there’s nothing special about the human, that it can be completely replaced by the machine, that we really are just blood and guts. I think that’s wrong, and I think you’re right – we have dreams, we have a subconscious, we have emotions. So it comes back to something like what the Pope is saying. A moral standpoint. We’re just not going to do this. Just because we can doesn’t mean we should. Which means shifting the incentives. And helping people actually come into an analogue life, because right now our society rewards people who don’t live analogue, slow lives. You have to be a very productive economic unit.

AG: It’s really hard not to be colonised by that type of thinking, or to virtue signal when someone says how are you – I’m so busy. I think I was doing that in this conversation. It’s some kind of contemporary micro value, and it’s so deeply embedded. And then here comes a tool that – oh wow, this is going to serve our productivity. Really difficult to resist when all of your settings are configured towards maximising productivity.

JK: And the lie is that you’ll do less with AI. I’m just doing more.

AG: In the same way – and I mention this in the essay – that the advent of the washing machine supposedly would liberate women from the laundry. But no, up it goes in terms of what constitutes an appropriate amount of domestic labour.

JK: Which is why I think we need a cultural renaissance in Australia. We’re in a vapid void right now and it’s being filled by tech instead of truth, beauty and goodness.

AG: I’m all for the cultural renaissance. Good idea. Bring it.



Sign up to analogue, a fortnightly newsletter for Australians (and friends elsewhere) who want AI to go well.