Politicians, regulators and technology leaders gathered in Dublin yesterday for the AI Summit, one of the landmark events of the Irish presidency of the Council of the EU, where they heard the Commission’s tech chief, Henna Virkkunen, set out a vision for Europe, and executives from Google, Microsoft and others make their case.
It is a sign of how far AI has moved from the margins to the centre of political debate that Micheál Martin will chair a new Cabinet committee on the subject, covering its effects on society and work as well as its regulation and use in the public service.
The questions raised by the acceleration of the technology are real, serious and potentially existential. For Europe, the prospect of being left behind once more is particularly pressing after years in which the EU’s economic growth has stalled and slipped ever further behind the US and China.
Then there is the looming prospect of mass job displacement and the social dislocation that could follow in its wake. Meanwhile, the industry’s seemingly insatiable appetite for ever more computing power is driving a data centre construction boom that strains local resources and is sparking opposition in affected communities.
READ MORE
Above all there is the prospect, promoted most vociferously by the AI companies themselves, that these systems are advancing so rapidly that they may soon escape human control and threaten the survival of our species.
All of this is far beyond the capacity of a one-day summit to address, but some things are clear. This is an industry that warns of catastrophe while asking governments for rules, though often on terms of its own choosing. Europe, for its part, is delaying some of the high-risk obligations in its own AI Act until 2027 and 2028.
One part of this complex jigsaw is already the subject of intense legal and political argument. A parallel conference in Dublin yesterday, held by the news publishing and recorded music sectors, represented groups experiencing the real consequences of displacement and disruption. To train their models, AI companies have used vast quantities of copyright material produced by authors, journalists, composers and other creators. One company, Anthropic, has agreed to pay $1.5 billion to settle claims over pirated books, without admitting wrongdoing. OpenAI and Microsoft await the outcome of their court battle with The New York Times. AI-generated answers are replacing clicks on search results, depriving websites of traffic.
These are important questions for the survival of the creative industries but also for how we define the future relationship between human capabilities and these immensely powerful, privately-owned machines.
