Breda O'Brien
Recently, I was sent a video of Donald Trump tenderly cutting Kamala Harris’ hair. It was obviously a fake but good enough to cause a quick giggle. Reading last August that a private school in London had instituted the first fully AI classroom was not quite so funny.
David Game College charges 20 young people studying for their GCSEs £27,000 each for the privilege of a fully automated learning process allegedly tailored to each student’s needs.
It was not quite as teacher-free as it might appear. Three coaches have been employed to maintain order and to teach subjects like sex education and art, where a human presence is felt necessary.
One of the so-called coaches is a former Latin teacher. What does it say about our culture that someone has moved from teaching the classics to being an assistant in an AI classroom?
Most parents will be unable to afford fees like those for David Game College but the major tech corporations such as Google and Microsoft are already embedded in schools from primary school onwards. Their GenAI applications are already being pushed in schools.
The Covid-19 pandemic exposed a digital divide between students, where poorer students were trying to access classes and homework on a mobile phone while better-off students had a laptop or tablet. The AI divide threatens to eclipse the digital divide. Those on higher incomes will be able to afford more sophisticated AI.
Controversy
In Ireland, much of the controversy around AI in education has centred on the proposal to allocate 40% of the total marks in each subject in the Leaving Cert to coursework. There is nothing particularly new about coursework. Many practical subjects from music to technical graphics have had it for years, along with some of the humanities.
The problem is that now the coursework will have to be designed so that it cannot be AI-generated and frankly, no one knows how to do that. What works today will be superseded by technological development tomorrow. Humphrey Jones, chair of the Irish Science Teachers’ Association, generated a H1 answer to a sample coursework question in moments using AI.
While this may be the most urgent question, given that the reformed science subjects are being rolled out in August 2025, it does not begin to address the questions posed by generative AI not just for education but for society.
The current generation of AIs has been trained on massive data sets scraped from the internet without permission”
The EU has gamely tried to keep on top of developments with its AI Act, which proposes seven principles for ethical AI usage, including human oversight and agency. Given that we do not even know fully how Generative AI reaches its conclusions, this seems more aspirational than realisable.
The same is true of other aims, such as transparency, and privacy. The current generation of AIs has been trained on massive data sets scraped from the internet without permission. Some large institutions like the Wall Street Journal and the New York Times have started to fight back, demanding compensation for the use of their data.
That is all well and good, but what about those who cannot launch massive lawsuits? I think in particular of graphic designers. Numerous AI apps brazenly suggest that you upload an image and the app will produce a creative variant, in other words, it will steal the original artwork and alter it.
Copyright was always a problem for artists but now it will be impossible to enforce.
When I did a Master’s Degree 20 years ago, I engaged a human being to transcribe the many interviews I conducted. Today, I would just feed the file into an AI app and receive high-quality transcripts. It means that someone who worked part-time from home to generate extra income is out of work.
Fulfilling
You could argue that it was never particularly fulfilling work but many jobs that people currently find fulfilling will also be affected. It is not just graphic artists but also musicians and writers.
It is deeply unjust that people’s work has been used to train AIs that will replace them. At the moment, AI is prone to hallucinations, some of which are difficult for humans to comprehend. For example, it frequently gets the number of fingers on someone’s hand wrong.
To a human being, this seems crazy but GenAI is just a sophisticated pattern decoder and re-creator.
It does not know that seven-fingered humans are not the norm. However, such problems will eventually be ironed out.
A world where huge tech corporations push us into fewer and fewer human interactions and more and more bland, derivative art and music, is a far bigger problem. To date, I have not seen evidence of any strategies to address it at the profound level of its impact on human flourishing. It is as if our imaginations refuse to acknowledge that a revolution greater than Gutenberg is already underway.