I’m on sabbatical for the next three months and have committed to doing a literature review on neurodiversity and PhD study. Ugh. I hate doing literature reviews. I’m just going to say it: most academic writing is BORING and doing a big review means reading lots of it. The thought of reading more than 200 papers in 3 months makes me want to poke my eyes out with a pen.
The good news is I can use ChatGPT to help me write this review. There’s something of a moral panic around ChatGPT, or, as my twin sister amusingly calls it: ‘ChattieG’. We’ve heard much about the perils of misinformation and how to stop students cheating. While there are legitimate concerns about ChattieG around privacy, pollution and worker exploitation, I think a lot of the commentary misses the point.
Look, here’s my hot take, for what it’s worth: academics have a responsibility to show others, including our students, how to use these tools responsibly, creatively and ethically. Feeling conflicted about using the technology is reasonable because there is no ethical consumption under late stage billionaire-ist capitalism. While we should continue to critique how this technology is made and used, refusing to engage is not really an option. The genie is well and truly out of the bottle and the robots really will take our jobs if we don’t put ourselves in charge of this technology.
Ok, I will get off my high horse now.
It took me quite some time to work out how to make ChattieG genuinely useful. The rest of this post will document what I have learned so far, but check with your uni before you follow my lead. Here’s the advice from ANU, which is kind of vague and allows for exploration, unlike some others.
Generally, the best way to use ChattieG is to imagine it as a talented, but easily misled, intern/research assistant who has a sad tendency to be sexist, racist and other kinds of ‘isms’.
Instruct it like you would a human being with these shortcomings and you will get much better results. And I know this sounds weird, but it also helps give ChattieG an actual face. I typed in ‘cute, smart intern wearing glasses’ into AI image generator DALLE, and this is what I got:
Of course, ‘cute and smart’ means young white guy (sigh). But I suppose the image is a good reminder of the bias that seems horribly built into AI tools of all stripes. Ok, to the literature review – how can it help?
You can ask ChattieG to suggest ideas. While ChattieG can be a bullshit fountain, the bigger problem is its ideas are likely to be boringly pedestrian. Boring is not all bad: it can be a starting point for fact checking or to engage in some myth busting.
Let’s see how it does with my basic project idea, which is to help supervisors and universities better support neurodiverse doctoral students:
Prompt: I’m writing a literature review about neurodiversity in doctoral education. Can you list ten things that educators should know about neurodiversity when supervising doctoral students? Please touch on aspects of neurodiversity that can affect how people write, process information, approach deadlines and communicate with others. include a short description with each list item, written in academic style.
Response: Certainly, here are ten key things that educators should know about neurodiversity when supervising doctoral students:
- Neurodiversity refers to the natural variation in how human brains process information and perceive the world around us. This includes individuals with conditions such as ADHD, autism spectrum disorder, dyslexia, and other learning differences.
- Neurodivergent individuals often have unique strengths and abilities that can be leveraged to their advantage in academic settings. For example, some individuals with ADHD have been shown to be highly creative and innovative problem-solvers.
- Neurodivergent individuals may face challenges with executive functioning, which can affect their ability to plan, prioritize, and meet deadlines. Educators should be aware of this and work with students to provide support and accommodations as needed.
- Many neurodivergent individuals have different communication styles and may struggle with social interactions. Educators should be patient, flexible, and willing to adapt their communication strategies to meet the needs of their students.
- Neurodivergent individuals may process information differently than their peers, which can impact their ability to write and articulate their ideas clearly. Educators should work with students to identify strategies for organizing and expressing their thoughts effectively.
- Neurodivergent individuals may benefit from alternative forms of assessment, such as presentations or projects, that allow them to showcase their strengths in non-traditional ways.
- Educators should be aware of the potential for bias and discrimination against neurodivergent individuals in academic settings. They should work to create a welcoming and inclusive environment that celebrates neurodiversity and values diverse perspectives.
- Accommodations such as extended deadlines, quiet study spaces, and assistive technology can make a significant difference for neurodivergent students. Educators should work with students to identify and implement these accommodations as needed.
- It is important for educators to recognize the intersectionality of neurodiversity with other identities such as race, gender, and sexuality. They should be aware of how these factors can impact the experiences of neurodivergent individuals in academic settings.
- Educators should prioritize self-education and ongoing professional development around neurodiversity to ensure that they are providing the best possible support for their students. This may involve attending trainings or workshops, consulting with experts in the field, or engaging in ongoing research and reading.
That’s… not bad. There’s some good ideas here for me to follow up. I particularly like the one on intersectionality, which prompted me to widen my search terms in useful ways.
Use ChattieG to make tiny summaries of key concepts
An important part of note taking is putting other people’s concepts and ideas in your own words. ChattieG can do this for you by simplifying other people’s writing; translating the academic concepts and making them easier to understand. I use this feature to make tiny summaries of interesting ideas from papers I’m reading.
If the text you want to simplify is on the web, you can just point Chattie at the URL and it will do the whole document. But I prefer to use this feature on smaller chunks of text as you get more accurate results. Here’s a bit of text from a paper I found after taking ChattieG’s advice to explore intersectionality: “The Theory Of Intersectionality: A New Lens For Understanding The Barriers Faced By Autistic Women.” by Amanda Saxe published in Canadian Journal of Disability Studies in 2017. I understand these concepts, but the writing is very dense:
“Intersectionality takes into account that “identities are …culturally mediated constructs implicated in relations of power, privilege and oppression” (Liasidou, 2013, p.300). This is directly relevant to the social model of disability, conceptualized by Oliver (1983) in Social Work with Disabled People which proposes that disabilities are the result of societal barriers, rather than inherent deficiencies in the ‘disabled’ person (see also Appleby, 1994; Ballan, 2008; Hirschmann, 2012; Oliver, 1983). Individuals with disabilities face barriers constructed within the able-bodied communities in which they live (Ballan, 2008; Hirschmann, 2012; Oliver, 1983). They could lead ‘normal’ lives should these barriers (both physical and social) be removed, since it is these barriers that disable them (Hirschmann, 2012). Throughout history, Autism has been seen through a medicalized lens, where individuals with the disorder were viewed as deficient and abnormal (Bumiller, 2008). The neurodiversity movement has since emerged through the work of Autistic self-advocates, promoting the argument that Autism is the result of a difference in neurology which should be accepted as a form of human diversity”
Let’s see how ChattieG goes simplifying the key concept in this paragraph. You could just say ‘summarise’ and give it the text, but remember it’s better to talk to it like a research assistant. Tell Chattie what the source paper is about as well as reading level and style of output. I find instructing it to simply to a 9th grader level in academic style gives consistently good results:
Prompt: I’m going to give you some text from an academic paper about intersectionality and autism. Can you please produce a simplified version of this text. Make it to a 9th grader level, still in academic style: [text inserted here in quotation marks]
Response: Intersectionality is a way of looking at how different parts of a person’s identity, like their gender, race, and disability, can affect how they’re treated. People with disabilities have faced barriers to living a “normal” life because society isn’t set up to accommodate their needs. This idea is called the social model of disability. It says that disabilities are not caused by something that’s wrong with the person, but by the way society is set up. In the past, autism was thought of as a problem that needed to be fixed. But now, people with autism are saying that it’s just a different way of thinking and that this should be celebrated as a form of human diversity.
An important part of the note taking process is what I call ‘rambling’ – just typing out my thoughts for an audience of one: me. Rambling is a way of getting ideas out of my head and onto the screen so I can assess whether they are any good. Chattie is brilliant for taking these rambles and turning them into something… less shit.
Prompt: I’m going to give you some of my rough notes about how neurotypicals process information compared to people with ADHD. Can you rewrite these notes as the discussion section of a paper about the differences between how people with ADHD work, vs people with neurotypical functioning: “I’m going to give you my rough notes on what it’s like to work with ADHD. Can you turn these notes into an introduction to an academic paper? “I list each segment of the event in a table and then write the tasks associated with each part of the event. As I do this, I must imagine – literally walk through each part of the event in my mind. This is DMN in action, then I flip back to TPN to put the task into words on a page. Then I must imagine a future flow of work where I can do these tasks – which one needs to be done before the other, are there dependencies? I then flip to my diary and look at the work that is coming up – are there any ‘big rocks’ that are going to get in the way? Can I schedule this work or must I delegate it. I then have to imagine my co-worker’s flow of work and tasks. I have to imagine the physicality of doing something like a stocktake of our tea and coffee supplies and how that might fit into her schedule, which I see visualised in the team diary. When I look at my work this way I wonder how I EVER get things done. And yet I do. For me, this work is effortful, but not difficult. But to a person with ADHD with the faulty switch between these different cognitive processes, it’s really difficult. They can get ‘stuck’ in the past, not the future – remembering a previous event rather than imagining a future one. If something went wrong on that previous event, they may get stuck in rumination about it. Although they want to get on with their task list, it’s a conscious thing to try and switch out of DMN and back into TPN.”