28 June 2024
Isobel Eddyshaw, University of Exeter
One of my goals in attending UKSG – an experience I very much enjoyed and found to be immensely valuable – was to learn as much as I could about how generative AI was being discussed and viewed in the library and information sector. How might we be working to combat some of its key problems: bias, theft, privacy violations, and above all its immense effect on the environment?
As Siobhan Haimé noted in a fascinating talk on “demystifying AI”, generative AI is exceptionally water-intensive: one ChatGPT query, she told us, has roughly the same ecological impact as one shiny metal UKSG water bottle. As such, Haimé mentioned donating to environmental charities to help offset the cost of her work in AI. She also, in Clarivate’s plenary talk on their own newly developed generative AI tool, asked the speaker what considerations the company had made towards mitigating their environmental impact. I was impressed with the depth of Haimé’s knowledge and the attention she paid to detail. Clarivate’s answer to her question was, however, less encouraging – the speaker simply didn’t know. When an answer to the question was provided, after the talk had ended, it stated merely that the company worked with Amazon, Google and Microsoft as cloud providers, and that each of these companies have their own 2030 net zero commitments. This is not reassuring: Amazon, for example, has recently been removed from the Science Based Targets initiative’s list of climate-conscious companies, having retracted some of their previously stated sustainability goals and failed to provide sufficiently clear road maps for meeting their remaining targets. Meanwhile, the ongoing devastating impact of these technologies on our environment is well-known, as one Nature article, “The carbon impact of artificial intelligence” by Payal Dhar, cogently explains.
I worry that we have a pervasive and disturbing tendency to believe that naming a problem is the same as doing something to fix it. Acknowledging the environmental impact of technologies such as generative AI while at the same time failing to do anything of value to actually mitigate or combat it, is simply not sufficient. Climate change is not just one drawback that we must carefully weigh against the benefits of a given tool. A machine that spits out quaint short stories, helps you format your CV and lesson plans, and summarises articles for you, will not be a comfort in the face of total environmental collapse. It is of course entirely right to say that we cannot ignore AI – our students are already using it, and will continue to do so. Other sectors outside of our own are capitalising on it too, and we must engage or be left behind. But if we are to actively involve ourselves in this area, we must do so seriously, and responsibly. At present, I am not sure that we are quite there yet.
Of course AI was not the only subject that I came to UKSG to learn more about. As a new subject librarian (still mostly trying to get my head around systematic reviews), I felt that I could stand to be significantly more informed on Open Access and transformative agreements. Tuesday’s Plenary Session 2 on this topic, in particular the talk given by Chris Banks and Caren Milloy entitled “Are we there yet? A review of transitional agreements in the UK”, was absolutely fascinating. I was keen to learn more about the impact (or lack thereof) of transitional agreements on engendering open access research, and this talk gave me a lot to think about. I now feel significantly better equipped to discuss this with the researchers I work with and our Open Access team, and am in the process of reading the full JISC report to help me engage further.
Relatedly, I was fascinated to learn about the new UKRI-funded platform for scholarly research, Octopus. This platform, which allows researchers to publish their work fully and for free, and enables community-driven peer review and collaboration, shows great promise in helping to ensure research is accessible, transparent, and of good quality. Octopus is still in development and open to feedback, and I have already started sending links to my PhD friends to get their thoughts: we are all excited to see how this platform could develop, particularly for the humanities (as it is currently geared more towards STEM subjects).
Another real highlight of the conference for me was the DOAJ game, developed by Katrine Sundsbo. An expert in gamification, Sundsbo has devised an entertaining and informative session in which teams work together to solve a series of puzzles, all of which educate the players on the history and goals of the Directory of Open Access Journals. Activities such as these are a great way to get information to stick in your mind, and this session was also the ideal choice for 4pm on the Monday – jumping up and down and running around the room trying to crack codes was an excellent way of waking me back up after a long day!
The organisation of the conference was also excellent, and I particularly appreciated the fact that a designated quiet room had been included. As someone with sensory processing issues, having somewhere properly quiet to escape to was essential in keeping me functional, and enabled me to engage with the conference far more effectively than I would otherwise have been able to. The kindness and support of everyone at the conference was also amazing – turning up to the first-timers reception on Sunday evening was incredibly nerve-wracking for me – right up until a kind trustee marched me towards a friendly group and introduced me. For the rest of the conference, I then had a group of friendly faces to chat to and attend seminars and events with.
The social events were also brilliantly thought out. Holding the gala in the science centre was, quite frankly, a stroke of genius – what better way to start a conversation with some librarian you’ve nervously been admiring from afar than by saying “have you seen the display that lets you shoot ping-pong balls into a basketball hoop using air pressure”? It’s a slam-dunk of a conversational opening gambit.
All in all, UKSG as a first-time has been an excellent, if tiring, experience, and I’m immensely grateful to have had the opportunity to attend. Meeting such an engaged and passionate group of people was a privilege. I hope, in future, that we can continue to harness this dedication, enthusiasm and initiative to ensure that we meet challenges such as climate change effectively, seriously, and responsibly, and I have every faith that this is possible.