The use of AI has divided opinion more than almost any form of technology since the emergence of nuclear energy. The speed at which it has proved possible to develop the technology means we are wrestling now with key questions somewhat sooner than some may have imagined.
While some think only negatively about AI, in the manner of the Terminator films in which the SkyNet system becomes self-aware and declares war on humanity, or in Luddite terms as a destroyer of jobs, others see it as making life and work easier, such as when machine learning improves the analysis of medical scans to spot problems a human eye might miss.
These issues may be considered by those who are thinking of using AI technology for big data inventory management. Among such organizations could be UNESCO, the United Nations body responsible, among other things, for documenting a huge amount of information relating to its own World Heritage List.
Not only does UNESCO have to catalog 1,199 sites, but also monitors them, with some sites being classed at risk or even removed due to changes that erode the heritage assets that put them on the list.
An example of this would be the city of Liverpool, which was listed until 2021, not for being the home of the Beatles, but for its maritime heritage. UNESCO took a dim view of the redevelopment of its historic docklands. The Elbe Valley in Germany suffered the same fate in 2009.
While nobody is suggesting AI played a role in such controversial decisions, UNESCO has a wider concern about the ethics of AI and held a forum earlier this month to reflect on the issues involved. It stated that “getting AI governance right is one of the most consequential challenges of our time”.
Among the ethical concerns UNESCO raised were “the potential AI systems have to embed biases, contribute to climate degradation, threaten human rights and more,” all of which it argued were likely to further disadvantage those who are impoverished and marginalized.
All that might sound rather alarmist, especially for any organization that simply wants to collate its data more efficiently. Among them will be organisations that, like UNESCO, manage huge heritage databases and for which easier management of large amounts of data would make life easier.
Of course, AI can be used in all sorts of areas unrelated to things like heritage databases. While it may be great at helping diagnose diseases from scans, it might also spot issues that come up in routine medical tests that demonstrate a higher risk of getting a disease in the future, such as through genetic predisposition.
That kind of pre-diagnosis could be very useful for helping someone keep an eye out for early signs of a disease they are at high risk of getting, but the bigger ethical question could come in the impact on health or life insurance premiums. However, such issues already exist.
By contrast, when it comes to heritage inventories, big data assessments may help make fair and consistent decisions on the listing of heritage assets. The stakes may not be as high in such situations as in the scenarios UNESCO has raised the alarm over, but it will still matter when more big calls have to be made in future situations like those of Liverpool.
