There’s much discussion in some of the circles I inhabit about ChatGPT and similar technology. If you aren’t up on the details, these “ChatBots” are forms of artificial intelligence that generate writing designed to imitate human conversations. Depending on your stance, they are either the next big thing and will make life easier, or they are evil and will undermine the very fabric of society.
In the academic circles I mix in, there is great concern students will make use of ChatBots to complete assignments for them. Academics are anxious it will be difficult to distinguish original work written by students from that written by a ChatBot. In the entrepreneurial world of online course writers I also inhabit, there is excitement ChatBots will make life easier. For people who are not confident writers, being able to have a ChatBot generate the first draft of your course or your sales page is an attractive prospect.
I (mostly) enjoy writing and I’m pretty good at it, so I don’t feel the urge to have a ChatBot take over the generation of my blog posts or another other of my writing tasks. I’m also anxious about handing over my email address and phone number to random strangers on the internet, so I haven’t set up a ChatGPT account to play with it. But I am curious about what we might learn from ChatBots.
I’m certainly no expert on how they work, but it is my understanding ChatBots examine written information available online, collate it, and then use it to build new writing. As someone who is interested in and has used discourse analysis as a research tool, ChatBot generated writing seems to me to offer an interesting way to capture “the mood” of what is currently being written about any particular topic. When the online platform I am using to build and deliver online courses (Kajabi) announced they had introduced ChatBot technology, I was keen to play with it to see if it could indeed be used this way. Turns out you can!
What happened next?
I asked the integrated ChatBot in Kajabi to generate a lesson about the topic “what does research say about CTG use in labour” and this is what it came up with:
Cardiotocography (CTG) is a medical procedure used to monitor the fetal heart rate during labor. It is used to detect any changes in the fetal heart rate that could indicate a problem with the baby’s health. It is usually done in combination with other tests, such as ultrasound scans and fetal monitoring. The aim of CTG is to provide early detection of any potential problems so that appropriate medical intervention can be provided.
Research has shown that CTG is a useful tool for monitoring the fetal heart rate during labor. It can detect any changes in the fetal heart rate that could indicate a problem with the baby’s health. It can also help to identify any potential problems before they become serious. CTG can also be used to monitor the progress of labor and to provide information about the baby’s wellbeing.
Overall, research has shown that CTG is a useful tool for monitoring the fetal heart rate during labor. It can provide early detection of any potential problems so that appropriate medical intervention can be provided. It can also help to identify any potential problems before they become serious. CTG can also be used to monitor the progress of labor and to provide information about the baby’s wellbeing.
- CTG is a medical procedure used to monitor the fetal heart rate during labor.
- Research has shown that CTG is a useful tool for monitoring the fetal heart rate during labor.
- CTG can provide early detection of any potential problems so that appropriate medical intervention can be provided.
So how did it do?
The ChatBot has constructed proper sentences, so it reads like something a human could have written. It is very repetitive, but I guess that can sometimes be handy in a lesson where you really want to nail home certain points. There are inaccuracies in the third sentence of the first paragraph and the fourth sentence of the second paragraph: CTG monitoring is a type of fetal monitoring, rather than something it is used “in combination with”. And CTGs don’t provide information about labour progress.
It does resemble the type of information that is freely available on the internet for women about CTG monitoring. There’s an important aspect of the text generated by the ChatBot I want to draw your attention to, and it highlights concerns I have about other written sources of information for women.
Yes, CTG monitoring is a tool used to monitor the fetal heart rate in labour. So far so good. Yes, it is used to detect changes in the fetal heart rate that COULD indicate a problem. There’s no real fact claim here – the text is not saying it does indicate whether there is a problem or not. And that’s the problem with this and similar bits of writing about CTGs – the “facts” are rather slippery.
Yes, the AIM is to provide early detection of POTENTIAL problems. Again, there’s no real fact here. I could claim the piece of lapis lazuli I keep on my shelf aims to provide early detection of potential computer problems. That’s not possible of course, but it could still be my aim. (It isn’t though, I just like the look of it.) Note also how the problems only need to be POTENTIAL ones to justify an intervention, not “real” ones.
Does research show CTG monitoring is a USEFUL TOOL? Research shows it is a very frequently used tool. So much of intrapartum maternity care is shaped by data from the CTG so, yes, I guess it can be argued it is both widely used and useful. But what does this particular text say it is useful for?
The text says CTG monitoring is useful for the “baby’s wellbeing” or “baby’s health”. There’s a whole other discussion to be had about fetal personhood and how labelling the fetus a baby is problematic, but that’s for another day. But what really is “wellbeing” or “health” in relation to the fetus, and how do we measure it? Note again how the terminology here is vague and doesn’t really mean anything. It is widely recognised that CTG monitoring is a screening tool with a very high rate of false positives when it comes to picking fetuses who are at risk of death or permanent injury.
Note how it is difficult to find a fact claim to argue against in the text the ChatBot has generated. The “facts” in the text are all rather slippery… This slippery language makes it possible to write three paragraphs and a dot point list without really saying anything at all about what research really says about CTG monitoring. In this regard the text reminds me of the preamble to so many CTG monitoring courses, advice for pregnant women, and the conversations maternity professionals have with women when they are recommending CTG monitoring.
What isn’t there?
The real aim of CTG monitoring is to prevent damage to the fetus from hypoxia leading to death or long-term neurological damage. There’s no statement in this text about whether or not CTG monitoring achieves that aim. Nor is there any information about the impact on women who use CTG monitoring in their labour.
There IS research about both (potential fetal benefit and maternal harm) in relation to CTG use. But that’s not what the ChatBot gave me. It gave me vague nice sounding platitudes instead. This text reminds of being patted on the wrist and told “don’t you worry dear, you leave the thinking to us”. Remember, the reason the ChatBot generated this text is that it is typical of what exists out there on the inter webs. Which says a lot really….
As someone who conducts, and writes about, research relating to fetal heart rate monitoring, I’m not in the least bit concerned ChatBots are going to do me out of a job any time soon. What this exercise has done is show me the absolute necessity for continuing the work I do. Someone needs to be regularly putting facts out into circulation because it clearly isn’t being done well already!
Categories: CTG, Language, Reflections, Writing
Tags: Artificial intelligence, ChatBot, discourse analysis, Kajabi
Leave a Reply