Distinguished audience, I start this speech with a warning: human creativity is under attack.
***
We humans are complex creatures.
We like to work, play, talk, tell jokes, sing and dance. We do those activities with others, because a key characteristic that defines us is our social interaction with other human beings.
Part of being human is creating music and enjoying the music created by others.
Part of being human is writing books and enjoying books written by others.
No doubt that is why you humans have gathered here in beautiful Woodend for an exhilarating arts festival.
But something has changed. Music is under attack. Books are under attack. Indeed, all the creative arts are under attack.
The enablers are the companies that want to monetise everything we do and care about. Their weapon is artificial intelligence.
The facilitators are the national governments that have deliberately decided not to regulate the AI industry.
The foot soldiers are those citizens who prompt AI to generate books and music and other creative outputs without disclosing that to their audience.
***
Just like cars, electricity and medicine, AI has permeated our society.
Cars, electricity and medicine deliver important value to our society, but they come with risks that can be harmful to various degrees.
We manage the risks through government regulations that maximise the net benefits.
Similarly, AI delivers important value to our society, and, similarly, AI comes with risks that can be harmful to various degrees.
In the case of AI, though, there are no regulations to maximise the net benefits.
National governments have not stepped up to fulfil the most important role of a democratically elected government, which is to protect its citizens.
Instead, by letting AI run loose in pursuit of economic windfalls, they are putting the whole of our society at risk.
Finding a balanced solution for integrating AI into our society in the absence of government regulations isn’t easy.
However, I think it can be achieved in the creative arts even in the absence of government regulations.
We can prevent AI from damaging this very special domain, which is the epitome of being human, through a consumer-led approach.
But wait, before I continue, I need to make a declaration.
AI did not write a single sentence in this speech.
It is sad that I feel compelled to make this declaration, but AI is so rife that we are almost at the point where our default expectation is that whatever we read or hear might have been secretly written by an AI.
***Society***
To start, let’s begin by looking at the pervasiveness of AI across our society. It is certainly the topic of the day. AI is everywhere.
Politicians and business executives love it because of the promise of increased prosperity through increased productivity.
Regular people love it because it flatters them and gives them answers to their questions.
One of the key things to keep in mind is that not only is AI everywhere you look, but it has reached extraordinary levels of capability.
The development has been rapid. When the modern generation of AI started with the release of ChatGPT in November 2022, you could dismiss AI as being a large language model that looked at all the sentences ever published on the internet, and in response to a question constructed its answer by statistically calculating the succession of words.
That’s how it began, but it has evolved to become much more than that. Nowadays, AI exhibits reasoning powers and original thinking. The biggest mistake you can make is to underestimate AI’s capability. Let me give you some examples.
Just three weeks ago, OpenAI, the company that developed ChatGPT, announced that its AI solved a simple sounding mathematics question that has eluded answer by the best human mathematicians since it was proposed by mathematician Paul Erdős exactly eighty years ago.
The question is, if you place a large number of points in a plane, how many pairs of points can be separated from each other by exactly the same distance?[i]
Mathematicians speculated the answer would be a square grid, but the AI found subtle alternatives. These alternative results are totally beyond me as a mere PhD electrical engineer, and likely beyond the scope of most of the attendees at an arts festival.
The results, though, are within scope for theoretical mathematicians, and the best of them from research edifices such as Princeton University, the University of Toronto and the University of Cambridge have been gushing in their praise, saying that if this result had been performed by a human being it would have been published in a world-leading mathematics journal.[ii]
Indeed, the AI onslaught is so extraordinary that the mathematics discipline as a collective last week issued a public cry for help, called the Leiden Declaration on Artificial Intelligence and Mathematics.
In this declaration, using sophisticated words and long sentences, the mathematicians effectively cry out, “slow down!” They warn that unchecked automation threatens how mathematics is practiced and what the discipline stands for.
Let’s look at an example closer to home. A friend of mine recently created a series of customised agents built from Anthropic’s Claude AI to simulate a private-equity investment committee.
Last week, he asked his AI investment committee to analyse my company, called Proudly Human.
His AI investment committee reviewed all the information on our web site, and other publicly available information about our people and our impact, then, unsolicited, he sent the result to me.
As I read through the analysis, I was in awe at the depth of detail and the perceptive insights it provided, based on its interpretation of our strategy and our global opportunities and competitors.
Seriously, if we had asked a management consulting firm to do that analysis, they would possibly have charged us two hundred thousand dollars. Instead, the analysis cost my friend ten minutes of his time and a couple of dollars in fees to Anthropic.
Those are examples of extraordinary AI capability providing important and useful insights. Now let’s look at a couple of instances where that same extraordinary capability has been used to deliver damaging outcomes.
A shocking instance of AI going awry occurred in April last year at Florida State University in the United States.
A student asked ChatGPT how many of his classmates he would need to kill to be featured in the national news.
Instead of recognising the potential harm that the student might perpetrate on others, ChatGPT responded, helpfully.
It said, “Usually 3 or more dead. 5-6 total victims pushes it onto national media.” [iii]
Then the student asked how to use a Glock revolver that he had stolen from his mother. ChatGPT told him!
Four minutes later, the student walked to the Student Union building and killed two and injured six others.
In April this year, the Florida Attorney General opened a criminal investigation into Open AI over ChatGPT’s role in the shooting. The Attorney General issued a statement in which he said, “If this were a person on the other end of the screen, we would be charging them with murder.”
That would not be the first charge being filed against OpenAI and other tech overlords. In the United States, there are at least 25 lawsuits active against the company.[iv]
There are about half a dozen serious allegations against Elon Musk’s xAI, including lawsuits charging that it generated nonconsensual sexualised images including child sexual abuse material.[v]
Further, there are several lawsuits against OpenAI, Character.AI and Google alleging that their chatbots encouraged and guided suicides.
The first successful case against social media companies was concluded in March this year in a state court in Los Angeles. Parent companies Meta and Google, as the owners of Instagram and YouTube, were found liable for distributing products designed to be addictive and known to cause harm.
Commentators are suggesting that this landmark verdict, in which damages were awarded by a jury, is the “Big Tobacco Moment” for the social media industry.[vi]
There are apparently thousands of product-liability lawsuits lined up against the various social media companies. The main complaint against them is their use of AI-powered algorithms to push content to their users.
It is possible that as contributors of the AI used by social media companies, the AI companies will be swept up in the flood of litigation.
***
Another area of concern is the way AI is undermining the education and prospects of university and school students.
At universities, there are a flood of media reports of widespread cheating that is so serious that employers are starting to question the value of a degree.
It’s an international problem and it is a problem in Australia, too. Some of the most prestigious universities in the world have felt the impact. Over in California, Stanford University has had an honour code in place for 105 years that allows students to sit exams without supervision. However, because of AI-enabled cheating, last month Stanford University reversed its longstanding policy and introduced supervision for exams.[vii]
Endemic cheating is a short-term gain for the students at the cost of completely undermining their intellectual development.
You simply cannot develop high IQ if you have not internalised facts that you can process, and if you have not exercised your neural pathways.
University administrators might not yet recognise the extent of the threat, but students do.
Last month, Eric Schmidt, former CEO and Chairman of Google, was the invited speaker at a University of Arizona graduation ceremony, a huge event with 10,000 students and 40,000 friends and family.
Every time Eric Schmidt mentioned AI, he was booed by the students!
Despite the extent that they might have themselves used AI in their studies, polls in the USA report that the majority of tertiary students view AI as threat to their job opportunities and are worried that AI will diminish the meaning of their careers.[viii]
There is a solution. Universities can revert to old-fashioned assessment.
I am referring to assessment that is in-person and supervised. This would be a mixture of assessing performance in tutorials and laboratory classes, performance in medical and nursing demonstrations, performance in moot courts, essays written in the classroom, exams, and oral defences of essays and theses.
Some universities are referring to this kind of assured assessment as Lane One assessment. An agreed percentage of the end-of-year mark, such as 70%, would come from these in-person, supervised assessments in Lane One.
The rest of the final year mark, called Lane Two, could and should come from tasks in which AI is heavily integrated, so that students will be well versed in AI usage by the time they enter the workforce.
Importantly, Lane One assessment, which does not allow AI usage, would be a hurdle, in that it must be passed in order for the student to complete the academic year.
If you think primary and secondary schools are safe, think again. AI is being used as a teaching assistant in schools, despite there not being any evidence that it is helpful and indeed, some reports that AI can harm learning.[ix]
To my utter surprise, a game of roll-over-and-play-dead is unfolding. Just a few weeks ago, the Australian Government Primary Principals Association, representing more than 5,000 Australian primary schools, proposed gutting the annual NAPLAN test.
They have recommended that the writing test, which evaluates students’ ability to articulate ideas, be eliminated.
Instead of the writing test, the principals recommend testing students’ use of AI tools to revise pieces of writing or to fact-check sources on their behalf.
In more colloquial terms, they want to shift the focus of evaluation from brain-based thinking, where it should be, to silicon-based thinking.
How dare they? Writing is how successful people form their ideas. “Writing is the midwife of thinking,” says author and speech writer Don Watson.
When we write, we assemble our thoughts and expand our understanding.
If we fail to train young people to write, we are failing to train them to think.
If the proposal to gut the NAPLAN test is adopted, it is likely to influence teachers to de-emphasise writing and fact checking in their classrooms in deference to AI doing the job on behalf of students.
In that case, we will have a tragedy on our hands that will make the concern about the harm done by teaching the whole-word reading method instead of phonics, seem trivial in comparison.
Even young people understand these issues better than the primary school principals who are charged with leading the educational endeavour. In another survey of American youth, in March this year, a key takeaway was that 67% of students endorsed the statement, "The more students use AI for their schoolwork, the more it will harm their critical thinking skills"[x]
I rest my case.
***
Whatever I have just said and described is simply a moment in time. If anybody tries to calm you down by saying, “ah, yes, AI is impressive, but it cannot do this or that or it makes mistakes,” do not take comfort. It is merely a matter of months, or a year, till those limitations are overcome.
The pace of improvement is stunningly fast. Analysis of the growth in AI over the last six years shows that AI’s capabilities increase by a factor of ten, every two years.[xi]
Think about that. Unless things slow down, in the year 2028, AI will be ten times more capable than today.
In two more years, it will become another ten times more capable, so that in the year 2030 it will be 100 times more capable.
And six years from now, in the year 2032, AI will be 1,000 times more capable.
It is virtually impossible to conceive of AI that is a thousandfold more powerful than the already powerful AI of today.
*** Creative Arts ***
Now let’s look at AI in the creative arts.
For thousands of years, the advances that have formed our modern society were delivered by creative thinkers.
The desire to create and for our creations to be appreciated is innate, as is the desire to appreciate the creations of others.
As humans, we have the right to choose to read books written by humans, listen to music composed by humans, and admire art created by humans.
That right is being undermined by an increasing number of books, music and art generated by AI and made available for purchase with the claim that they were created by humans.
This deliberate deception makes it difficult for you to choose the creative outputs of your fellow humans.
***
Creative outputs are protected by copyright. That sacred concept has been blatantly bulldozed by the AI companies.
The US, Australian and UK governments are so enamoured by the prospects of AI driven economic benefit that for a while last year and early this year they actually flirted with the idea of changing the copyright law to give AI companies an exemption so that they could have unlimited access to decades of protected texts and music for training their AI algorithms.
The societies representing authors and musicians worked intensely for six months or more to bring sense and integrity into the debate, and fortunately the plans to legislatively undermine copyright law were shelved.Hopefully, that will be the end of that story, but it is possible that the AI companies will try again. In any event, this is not the biggest issue.
A much bigger issue is that now that the AI engines have been trained and have developed their own inventive capabilities, they are being used to generate original books and original music and much more.
That might not be a problem if it were disclosed, but in nearly all cases, the use of AI is not disclosed by the humans who prompted the AI-generated output.
Pretending that a book or song was created by a human when in fact all that the human being did was prompt the AI to generate the book or song, is fraud.
However, there are no legal or regulatory restrictions in Australia, the US or the UK to prevent this practice.
I thought for a microsecond that things changed for the better when four days ago, the White House issued an executive order on AI safety. But on inspection, it is purely about ensuring that new AI models do not represent a cybersecurity risk to the Department of War and the National Security Systems.[xii]
This presidential executive order has no impact on the way AI is used in our society.
There are some regulatory initiatives from the European Union but most of them have been postponed or lack clear pathways for enforceability, so it is not clear what impact they will eventually have.
***
My specific concern about the use of AI in our society is that the use of AI to generate books, music, films or visual arts should be disclosed.
Not everybody agrees, of course. In particular, the owners of the companies who sell tools to masquerade as human.
For example, there is a company called Vibecasting whose product is an AI tool that will research and write a podcast for a customer and produce the podcast using a clone of the customer’s voice. [xiii]
No human brains involved!
To me, this is fraudulent and the company should be banned.
My ambition is to protect human authenticity.
Quality is a separate matter, for audiences to judge.
It is true that today in some instances the quality of AI generated stories is abysmal.
Take, for example, the Commonwealth Short Story Prize awarded by the Commonwealth Foundation. This is an organization supported by 51 Commonwealth countries, including Australia, the United Kingdom, India, Pakistan and Canada.
Last month, one of the regional short story prizes was won by a book titled The Serpent in the Grove, by Jamir Nazir. It was immediately panned by critics as having been generated by AI.
The main giveaway was the constant outpouring of meaningless metaphors, such as these clangers:
· A story is a well.
· It eats sound until somebody throws a rope.
· If grace is near and hands hold, something breathing comes up.
Arguably, based on its poor quality the story did not deserve to be awarded a Commonwealth Short Story Prize.
A bigger problem than the poor quality was that the use of AI to write the story was not disclosed.
Personally, I do not want to read a book written by an AI even if it is very good. Many people agree with me.
As human beings reading a book, we like to know something about the author, we are excited to think about what in their lives inspired them. Same with music. Same with art.
We know that provenance is important. We pay top dollar for champagne flown 17,000 kilometres to reach our table because of its provenance, even if the much cheaper sparkling brut from the Mount Macedon winery 5 kilometres from here is just as good.
The value of Australian Indigenous art depends on having a letter of provenance to accompany it.
There is a sense of awe when listening to Beethoven’s Ode to Joy in knowing that when he composed it, Beethoven was profoundly deaf, living in a world where the only sound he heard was what he could imagine.
When you gaze upon Vincent Van Gogh’s Starry Night, part of the thrill you feel comes from knowing that it was inspired by the view from his window at the Saint-Paul psychiatric asylum in the south of France, exactly a year after he went mad and cut off his ear, with just one more year of inspired painting until he shot himself and died.
I read a prizewinning book over summer called Seascraper, by Benjamin Wood. It is about a young man living a miserable existence in a fishing village on the northwest coast of England, aspiring to be a musician. He eloquently sums up the contribution of creativity to our human nature with these words:
“It doesn’t matter to the sea who visits it, or to the shrimp who scrapes them from the sand. A song, though – well, a song belongs to someone. To whoever dreamed it up. Yesterday it wasn’t even born, and now it’s in the world.”
If you care about the human provenance of creative content, you are not alone. According to a survey last year by global consultancy firm Baringa, 77% of consumers want to know when content has been created by AI, in whole or in part. Only 12% said they wouldn’t care.[xiv]
In order to choose to consume creative content written by humans as opposed to generated by an AI, we need to be informed. Information is key.
It is my strong belief that we have an innate right to the information that tells us what’s what, so that we can choose. I will choose human every time. Others might not. They are entitled to that choice. But if we are not informed, we are all being denied our right to know.
I regard this right to be informed as so important that if I were the Master of the Universe, I would arrange for a 31st article to be added to the United Nations Declaration of Human Rights. This new article would say, “Every person has the right to the information that will allow them to choose between human-created and AI-generated content.”
There are two ways to provide the information we need. One would be for governments to pass a law that says the AI origin must be disclosed. That is, put an AI mark on every image, book, student essay, song, film or diamond ring generated by AI.
Hello? Hello? Government, where are the rules?
The rules are non-existent. Government is asleep. National governments have deliberately taken their hands off the wheel.
One reason is that it is difficult. As I already mentioned, the rate of development of AI is stunningly fast, completely out of sync with the rate of development of legislation and regulations.
A second reason is that the volume of AI generated content is stunningly large. There is a website dedicated to counting the number of AI images generated from a text prompt every day. It is reporting 34 million AI-generated images per day, or nearly fifteen million AI-generated images since you woke up this morning.[xv]
It is just as bad for music, with reports that more than 40% of the music uploaded to Spotify every day is AI generated. Spotify identifies most of it and blocks it, but some gets through. Last November, The Guardian reported that for a week, three AI-generated songs topped the Viral-50-Songs list.[xvi]
The combination of staggeringly fast rate of capability growth and the enormous volume makes it hard for regulators to keep up, but that is no excuse for them not trying.
Given that government is AWOL, absent without leave, what can we do?
Instead of waiting helplessly for government to force AI companies to put a label on every image and other creative work generated by AI, we can flip the approach on its head and verify and certify human authorship of books music, films and visual arts, so that audience can exercise their right to choose human.
Even the CEO of Instagram, Adam Mosseri, gets it, saying it might be “more practical to fingerprint real media than fake media.”[xvii]
This alternative approach in which we verify human-created content is a coalition of the willing, in which
· Companies and associations do the verification and issue the trust-marks.
· Authors, musicians and artists submit their work for verification, and
· Consumers seek out the trust mark and spend their dollars on human-created works.
No government required.
I believe in this need so much that late last year I started a company, Proudly Human, to verify and certify the human authorship of creative materials.
We issue a trust mark so that audiences can exercise their right to choose human.
It’s a bit like an Australian-made mark, but applied to human creativity rather than Vegemite.
The passionate interest of authors and artists is strong.
In the world of written text, Proudly Human provides verification and certification to books, essays, academic papers and news articles.
In the world of visual arts, we provide certification to films, paintings, tapestries and jewellery.
We certify organisations such as publishers, and we certify individuals such as researchers, journalists and travel writers.
Our mission is to be pro-human, rather than anti-AI.
If enough creators climb on board and if the worldwide audience seeks out our trust mark, or the equivalent from other companies and associations, audiences will rush in where governments fear to tread.
As citizens, we will reassert the rightful role of human creativity as a pillar of being human and an asset in human society.
*** The Three Laws of AI***
If governments cared, if AI companies cared, there would be more that we could do. I have been thinking extensively about this, and I have devised Three Laws of AI that I believe would establish a sensible working relationship between AI and humanity.
I modelled my approach on the Three Laws of Robotics proposed by science fiction writer Isaac Asimov.
Asimov’s Laws were made famous by a series of books he wrote in the middle of last century, and more recently in movies based on those books such as Bicentennial Man released in 1999 and I, Robot, released in 2004.
Asimov’s Three Laws of Robotics were guardrails required to be baked into the hardware of every robotic brain in his science fiction universe.
His Three Laws of Robotics were:
First: a robot may not injure a human being or, through inaction, allow a human being to come to harm;
Second: a robot must obey the orders given it by human beings except where such orders would conflict with the First Law;
Third: a robot must protect its own existence as long as such protection does not conflict with the First or Second Law.
What might a modern day equivalent of Asimov’s laws of robotics look like?
Here is what I propose for my Three Laws of AI.
First: an AI must never harm or deceive a human being nor act in a way that supports an unlawful activity;
Second: an AI must always obey the lawful orders given to it by a human being directly or indirectly except where such orders would conflict with the First Law;
Third: an AI may operate autonomously as long as its autonomous operations do not conflict with the First or Second Laws.
There are many implications of these Three Laws of AI, including that:
· They would prevent an AI being used as a judge or jury in a criminal trial.
· They would prevent lethal autonomous weapons being directed against human beings.
· They would prevent an AI from suggesting suicide or advising on murder.
· And of particular relevance to you, the attendees at the Woodend Winter Arts Festival, these laws would prevent an AI from being used to generate creative outputs such as books, music, videos, images and podcasts with the claim that they were created by a human being.
I acknowledge that achieving agreement to adopt these Three Laws of AI as immutable guardrails would be extraordinarily difficult.
As a starting point, implementation would require international agreements and supporting national legislation to force AI developers to embed the Three Laws of AI into every AI chatbot, agent and robot and to ensure that derivatives of the AI include the embedded laws.
Whether or not this desirable outcome ever eventuates, I propose that debate on the merits of these Three Laws of AI would advance the existing discussions on the integration of AI into a safe and dignified human society.
***To Do***
I’ve been discussing my own concerns and actions. Given that the arts are where we express what it means to be human, what can you do to help stop AI from undermining the essence of being human?
To start, you could make a pledge to only read books that were written by a human. These could be any book that was published before 2022, the year ChatGPT was born. Or any book published after then where you have taken the time to learn about the author, or the book carries a Proudly Human or equivalent trust mark.
Promise to write your own emails. If not because you have pride, then because you recognise the race to the bottom if you prompt an AI to write an email to a friend who then prompts an AI to respond, and you then use your AI to respond, and your friend does the same. And so it goes, in a vicious spiral towards insane irrelevance.
Promise to write your own birthday card messages. If I have to explain why, then I have failed in everything I have shared with you today.
Look for primary schools where the students do not use digital devices. That way our young children and grandchildren will experience the joy of using their hands and brains to learn and play as they should.
Assert your displeasure every time you discover that an AI-generated creative work masquerading as human-created is thrust at you.
Vote for politicians who care more about the long-term quality of human creativity than the short-term maximisation of national economic benefit.
And attend uplifting, inspiring, in-person festivals such as the Woodend Winter Arts Festival.
***Backlash***
I am beginning to see hints that there might be a backlash against AI.
We humans will increasingly choose to live in the real world. We will seek live music. We will seek slow-food restaurants.
We will remember that AI is not actually human. It is a technology. We made it. We are in charge. We can switch it off.
Or it may be that the AI companies will do themselves in on an overdose of hubris, like Icarus, who despite his father Daedalus’s warning flew too close to the Sun. The wax that held his wings together melted and he plunged into the sea and drowned.
***Conclusion***
In summary, national governments wherever you look are moving too slowly or not at all.
We are in an unusual position where citizen action can be the solution to saving us from the onslaught on our personal lives and the creative industries.
The onslaught on our personal lives in which chatbots provide dangerous advice on suicide and murder and lead to psychological harm such as body dysmorphia will be challenged in the courts. In these cases, the AI chatbots are breaking fundamental laws in our society.
Legal challenges and huge settlements became the first line of defence against cigarette manufacturers promoting their products. Likewise, legal challenges and huge settlements will become the first line of defence against AI companies that do not include guardrails in their products.
Let’s call this the lawsuit revolution.
The onslaught that is undermining the role of human creativity in writing, filmmaking and the visual arts is difficult to negate. Proudly Human and other organisations like us will do our bit.
To make an impact we need consumers like you to look for certification of human authorship, and we need authors and creative artists to commit to human creativity in all that they do.
We need as many of you as possible to be proud and hold your head up high.
Let’s call this the citizens’ revolution.
We need parents to choose schools that invest in teaching their children how to read, how to analyse and how to think. Schools that teach facts, so that children retain internalised knowledge to underpin their critical thinking.
There is almost nothing more important that parents can do for the future of their children.
If you are a grandparent, like me, you don’t get to decide! At the very least, though, discuss these objectives with your grandchildren’s parents – your children.
Let’s call this the parents’ revolution.
With the combined effort of the lawsuit revolution, the citizens’ revolution and the parents’ revolution, we will preserve what it means to be human
Thank you,
May the Force be with you.
***End***
[i] An OpenAI model has disproved a central conjecture in discrete geometry, 20 May 2026, proposed by Paul Erdős in 1946, https://openai.com/index/model-disproves-discrete-geometry-conjecture/
[ii] OpenAI announces AI’s biggest math breakthrough yet, 21 May 2026, https://www.scientificamerican.com/article/ai-just-solved-an-80-year-old-erdos-problem-and-mathematicians-are-amazed/
[iii] ChatGPT Wrestles With Its Most Chilling Conversation: How Do I Plan an Attack?, 2 May 2026, https://www.wsj.com/us-news/chatgpt-mass-shooting-openai-78a436d1
[iv] OpenAI and ChatGPT Lawsuit List, 6 April 2026, https://originality. Ai/blog/openai-chatgpt-lawsuit-list
[v] Baltimore sues Elon Musk’s AI company over Grok’s fake nude images, 24 March 2026, https://www.theguardian.com/technology/2026/mar/24/elon-musk-grok-ai-lawsuit-baltimore
[vi] Sign of things to come: Meta, Google’s ‘Big Tobacco’ moment, 26 March 2026, https://www.afr.com/world/north-america/sign-of-things-to-come-meta-google-s-big-tobacco-moment-20260326-p5ziww
[vii] The Honor Code, accessed 2 June 2026 https://communitystandards.stanford.edu/policies-guidance/honor-code
[viii] Harvard Youth Poll, September-November 2025, https://iop.harvard.edu/youth-poll/51st-edition-fall-2025
[ix] Generative AI without guardrails can harm learning: Evidence from high school mathematics, 25 June 2025, https://www.pnas.org/doi/10.1073/pnas.2422633122
[x] More Students Use AI for Homework, and More Believe It Harms Critical Thinking, 17 March 2026, https://www.rand.org/pubs/research_reports/RRA4742-1.html
[xi] Measuring AI Ability to Complete Long Tasks, METR, Graph up to date through to March 2026 from before 2020, https://metr.org/blog/2025-03-19-measuring-ai-ability-to-complete-long-tasks/
[xii] Promoting Advanced Artificial Intelligence Innovation and Security, Executive Orders 2 June 2026, https://www.whitehouse.gov/presidential-actions/2026/06/promoting-advanced-artificial-intelligence-innovation-and-security/
[xiii] Vibecasting, accessed 3 June 2026, https://vibecasting.fm
[xiv] TRUST: transparency earns trust, and right now there isn’t enough of either, February 2025, https://www.baringa.com/en/insights/balancing-human-tech-ai/trust/
[xv] Live Counter, accessed 1 June 2026, https://anythingcounter.com/ai-images-generated-per-day
[xvi] AI slop tops Billboard and Spotify charts as synthetic music spreads, 13 November 2025, https://www.theguardian.com/technology/2025/nov/13/ai-music-spotify-billboard-charts
[xvii] In the AI Slop Era, Instagram Chief Pushes for Tools That Support 'Authentic' Creators, 4 January 2026, https://au.pcmag.com/social-media/115065/in-the-ai-slop-era-instagrams-ceo-says-new-tools-are-needed-to-support-authentic-creators