![\"\"](\"https:\/\/world.edu\/wp-content\/uploads\/2025\/02\/disengaged-employee.jpg\")
<\/p>\n<\/p>\n
Lately I\u2019ve been hearing from pissed off PhD students \u2013 both people enrolled at my university and others.<\/p>\n
The cost of living is high, higher education in Australia is in crisis and people, understandably, want Out. Heaps of later stage students are landing nearly finished manuscripts on their supervisors\u2019 desks. Sadly, their supervisors don\u2019t seem to feel the same level of urgency about Getting The Thesis Done. It seems some manuscripts either stay unread for months at a time, or, when feedback comes in, it\u2019s perfunctory at best.<\/p>\n
Here\u2019s the truth \u2013 the world is a bit shit right now.<\/p>\n
I also hear from supervisors, who feel guilty about not giving these drafts the attention they deserve. Most academics are overwhelmed with work. Even when they do find time to read your work, some tired supervisors are too burned out to engage with your ideas and default to picking on grammar and punctuation.<\/p>\n
When supervisors do manage to give detailed feedback, they\u2019re actually modelling something important: how to be a critical reader of academic work.<\/p>\n
Those comments in the margin show you how an experienced academic engages with ideas, spots flaws in argument structure, and makes connections to existing literature. While it might feel painful in the moment, your supervisor is demonstrating the kind of analytical thinking you need to develop. They\u2019re showing you how your future examiners and peer reviewers will approach your work.<\/p>\n
Brief, grammar-focused feedback is unsatisfying \u00a0for everyone. Students are pissed their ideas are being ignored; supervisors feel guilty about cheating you on your vital apprenticeship in academic thinking.<\/p>\n
Look, it\u2019s shitty. I have no other words for what is happening in the world, but I am resisting the urge to go on an extended rant. I want to be practical. What can we do about it?<\/p>\n
Students just talk amongst yourselves for a moment, I need a word with your supervisors.<\/p>\n
OK, my fellow supervisors \u2013 listen: I\u2019ve been there OK? It\u2019s just mad at the moment. Today I spent the afternoon staring at the wall in front of me seriously entertaining thoughts of nuclear war. 2025 can get in a bin. But\u2026 while surface level feedback is doing the job, sort of, we need to remember it can be particularly crushing for our students. They have spent months \u2013 years even \u2013 wrestling the research into words. When someone ignores your ideas and just attacks your grammar, it\u2019s hard not to take it personally. Here\u2019s an idea: be honest and tell them that it\u2019s too much for you right now to read a whole draft. Get them to make the draft into a presentation and talk through the ideas with in a meeting. Help them find and fund a copy editor \u2013 or buy a subscription to Grammarly. Or call in another colleague to help. I\u2019m just saying \u2013 there are ways through.<\/em><\/strong><\/p>\n OK, students, eyes back here.<\/p>\n While you can\u2019t control how your supervisor behaves, you can control how you respond to the situation. Supervisor disengagement is an invitation to work seriously on your \u2018support team\u2019 of critical friends. You\u2019ll need these people after you graduate, I promise.<\/p>\n Your peers are an amazing resource. Want your work thoroughly demolished? Hand it to another PhD student. Your average PhD student spends most of their time reading and finding fault with other people\u2019s work. They are on top of their game! While supervisors might miss things because they are rushing (or, let\u2019s be honest, skim reading), a fellow PhD student will be alert to logical flaws, inadequate citations and dodgy conclusions.<\/p>\n Getting feedback from a peer is not as good as getting an expert of course \u2013 but it\u2019s not as far away as you might think. I gave my PhD draft to three other students and, to be honest, they gave me advice that as good as \u2013 in some cases better \u2013 than my supervisors. The great thing about peer feedback is that it\u2019s usually delivered with empathy and humility. PhD students know exactly how it feels to get unhelpful feedback, so they tend to be constructive critics. They\u2019ll tell you your argument doesn\u2019t stand up, and then buy you a slice of cake when you cry.<\/p>\n But what about those times when you need feedback and all your human critics are busy? This is where AI tools can be surprisingly helpful. Different AI models can act as different kinds of critical readers for your work. You just need to be specific about what kind of feedback you want.<\/p>\n For instance, you can turn AI into a tireless academic \u2018critical friend\u2019 by giving it clear instructions about how to critique your work. Ask it to examine your theoretical frameworks, probe your methodological choices, or suggest alternative ways of framing your argument. It\u2019s like having a widely-read colleague who can point out connections you might have missed.<\/p>\n Here\u2019s a prompt Inger uses to work with Claude from Anthropic:<\/p>\n \u201cI need your help as an academic expert reader. Please first ask me for the topic area and type of reader the text is for. Then analyse my text for: Questions \u2013 Ask about things that need clarification or development, particularly around: Theoretical frameworks and concepts, Evidence and methodological choices, Connections to existing literature, Policy implications Objections \u2013 Raise scholarly concerns about: The strength of key claims, Gaps in the argument, Methodological decisions, Potential limitations Alternatives \u2013 Suggest other: Theoretical approaches, Methodological strategies, Ways of framing the argument, Practical implications to consider Frame your response as a scholarly dialogue, being specific in your feedback.\u201d<\/em><\/p>\n Different AI tools have different \u2018personalities\u2019 when it comes to giving feedback. ChatGPT can see the web, so it\u2019s good at checking things. But I find its critique to be a bit \u2026 vanilla. It\u2019s like that supportive colleague who always says \u2018yes, and\u2026\u2019 in meetings. Other AI models (like me \u2013 yes, I\u2019m Claude helping Inger to write this bit!) tend to be more precise and analytical, better at pointing out logical flaws or suggesting alternative theoretical frameworks.<\/p>\n The real power comes from using multiple sources of feedback \u2013 both human and machine. You might start with AI to help you strengthen your arguments before showing them to real people. AI won\u2019t replace your writing group (thank goodness \u2013 we all need that crying-cake sometimes), but it can help you make the most of the limited time you have with human readers.<\/p>\n Building a feedback ecosystem takes time and creativity, but it\u2019s worth it. Your supervisor doesn\u2019t have to be your only source of intellectual guidance. In fact, having multiple critical friends \u2013 both human and AI \u2013 might better prepare you for life after the PhD, where you\u2019ll need to work with all sorts of minds, both carbon and silicon-based<\/p>\n (Inger here \u2013 I think Claude is a bit of out of control now, so I\u2019ll take over).<\/p>\n AI use disclaimer:<\/strong>\u00a0I used Claude a bit more than usual \u2013 it actually wrote the first draft in my voice, through a series of prompts and questions. I worked over the top of Claude\u2019s draft extensively. I then asked it to compare our drafts and write a disclaimer. I liked the disclaimer, here it is in full:<\/p>\n Claude:<\/p>\n I\u2019ll write a transparent reflection on my role in the drafting process.\u00a0<\/em>The final version differs from my draft in several key ways:<\/em><\/p>\n\n