What better way to start this blog by stating candidly (duty of candour) that in preparing this article, I asked ChatGPT to write me a 500 word blog on, “The use of Artificial Intelligence (AI) in social care in the UK and how social care providers could use ChatGPT”.
The temptation to use it lock, stock and barrel as my blog was alarming, but I resisted. I have decided to use ChatGPT as a way to generate ideas and unblock writer’s block when I am writing articles, proposals and thought leadership pieces.
I will however copy one of the paragraphs it gave me in its very coherent, instructive and thought-provoking article. (note the US spelling which is standard)
ChatGPT can also be used to support social care providers in decision-making. By analyzing large volumes of data, AI can identify trends and patterns that can help care providers to make more informed decisions about service users’ care. For example, AI could be used to identify service users who are at risk of falls and recommend interventions to prevent falls from occurring.
Only last week I led a seminar on AI in social care, and I included a couple of tech suppliers in the presentation. In reviewing the content of the seminar and the content of the ChatGPT generated article, it was amazing to note that the article reflected much of what the tech providers said. For example, we have already seen the way that natural language processing using AI has the potential to give a voice to people who are not often heard. It is great to see that AI can be used to increase people’s independence, which, for me, is one of the underlying goals of what we do in social care.
No one, however, has addressed the issue of how a tool such as ChatGPT could be used in every day care situations: it could potentially write care plans for people, given the right information to input, and could be used to help people in social care formulate ideas. This is a subject I am really interested in. Moreover, at last there is a veritable competitor for Google, in that searches requested via ChatGPT do not appear to have any sponsorship content and give the facts as they see them. In a typical Google search, many of the top search items will be sponsored, which (personally) I find quite annoying, whereas the output of ChatGPT does not seem to have the same deficiency, although some people do detect its own information-led biases.
I would love to put together a compendium of how care providers are using AI, not just with tech solutions that they are implementing, but also in their everyday operations. I encourage you to try using ChatGPT to see how you can use it responsibly and with integrity, as I have tried to do.
I’m tempted again to use the concluding paragraph which was generated by the AI, but I will again resist. This is mainly because I think ChatGPT was biased in its positive portrayal of what AI can achieve in social care, but then maybe that’s my fault for not specifying carefully enough that I wanted an unbiased critique of the power of AI.
Please do let me know of your own experiences.
Daniel Casson is Care England’s Digital development advisor