Very selfishly I am delighted that AI filled your horizon Helen. Your writing on the wider context, considerations and consequences is an incredibly valuable reference and reflection point for me. Thank you
Thanks Helen for another great (if slightly depressing) post. Can you really see no positives for leaning and knowledge out of generative Al? I was a bit taken aback to hear you say this was so obviously a wrong idea.
Hi Antony, thanks for your comment. Of course some ppl are making good use of generative AI, particularly if they already have an established writing/coding/creative practice and secure knowledge of a field, and can discriminate which outputs are helpful. I think generation probably make it harder for novices to develop these, though the research is limited. I know there is good critical work going on with students. My sense is that it is mainly defensive. But my focus is on the impacts beyond immediate contexts of use, on knowledge ecologies such as publishing and teaching resources, and on the relationships between students and teachers/assessors more broadly. I am deeply sceptical of 'AI' as a narrative about the future of work, which I read as an apology for automation, precaritisation and restructuring. You don't have to share my critique of big tech's agenda to think that generative models are biased, discriminatory, exploitative, and dangerously unreliable. Where I am optimistic is in thinking the university sector will find a steady tone of scepticism and sound guidance, perhaps even of regulation. It's very hard to maintain that against the other narrative, though, which is why I try to be so consistent on the 'down' side.
Thanks for taking the time to answer my question, I see now that you were referring more to the second order effects rather than the direct implementation. It will be important (and hard) for universities to hold out against the aggressive marketing of AI tools via subscription from big tech and instead develop their own alternatives, as you mentioned in your 10 Minute Talk with Tim Fawns that I came across the other day. The issue right now is that one can only really critically evaluate the actual capabilities and value of tools by using them, and the best ones are all paid. Understanding those second order risks is also hard to fully appreciate without spending time with the tools or reading about them a lot, which not everyone has time to do. I suppose now that these models are everywhere the only way out is through!
Very selfishly I am delighted that AI filled your horizon Helen. Your writing on the wider context, considerations and consequences is an incredibly valuable reference and reflection point for me. Thank you
Thanks Helen for another great (if slightly depressing) post. Can you really see no positives for leaning and knowledge out of generative Al? I was a bit taken aback to hear you say this was so obviously a wrong idea.
Hi Antony, thanks for your comment. Of course some ppl are making good use of generative AI, particularly if they already have an established writing/coding/creative practice and secure knowledge of a field, and can discriminate which outputs are helpful. I think generation probably make it harder for novices to develop these, though the research is limited. I know there is good critical work going on with students. My sense is that it is mainly defensive. But my focus is on the impacts beyond immediate contexts of use, on knowledge ecologies such as publishing and teaching resources, and on the relationships between students and teachers/assessors more broadly. I am deeply sceptical of 'AI' as a narrative about the future of work, which I read as an apology for automation, precaritisation and restructuring. You don't have to share my critique of big tech's agenda to think that generative models are biased, discriminatory, exploitative, and dangerously unreliable. Where I am optimistic is in thinking the university sector will find a steady tone of scepticism and sound guidance, perhaps even of regulation. It's very hard to maintain that against the other narrative, though, which is why I try to be so consistent on the 'down' side.
Thanks for taking the time to answer my question, I see now that you were referring more to the second order effects rather than the direct implementation. It will be important (and hard) for universities to hold out against the aggressive marketing of AI tools via subscription from big tech and instead develop their own alternatives, as you mentioned in your 10 Minute Talk with Tim Fawns that I came across the other day. The issue right now is that one can only really critically evaluate the actual capabilities and value of tools by using them, and the best ones are all paid. Understanding those second order risks is also hard to fully appreciate without spending time with the tools or reading about them a lot, which not everyone has time to do. I suppose now that these models are everywhere the only way out is through!
Thanks for helping to make sense of this new world!