We know that making progress is a critical motivational factor in finishing a PhD and maintaining good mental health while we do it. In turn, our productivity plays a big role in whether we make progress on our dissertation or not. As the first post in a series on doctoral productivity, I could not help but fall into one of the thèmes du jour: whether ChatGPT and similar artificial intelligence (AI) tools can write a good piece about this topic. In this post, I go over a couple of iterations of (pretty good) computationally-generated advice, and finally give you a few ideas that I think are overlooked by the algorithm.
Productivity is one of the big topics we often cover in this blog (and in the workshops we run to support doctoral students1). As I was preparing a series of posts on doctoral productivity, I was also marinating in a lot of the new AI-related literature and applications (on which I am focusing in my own upcoming research work2). Also, I am seeing more and more often doctoral students making use of these AI tools (and not always in ethical or effective ways). Thus, lately I couldn’t help but wonder whether efforts like this blog are (or will soon become) obsolete, in the face of this new generation of AI tools. Shall a doctoral student now look for, e.g., productivity advice in my blog, or just ask for advice (even tailored for their particular situation) to ChatGPT? In the end, the algorithm was probably trained also with the content of my blog and Twitter feed, so some of my knowledge (and that of others smarter than me) may be hidden in the gazillion parameters of the algorithm. Thus, I setup a little fun experiment: see if can I get some good doctoral productivity advice out of ChatGPT, and whether I can write something “better” than the algorithm’s output. All this, spending more or less the same time I would spend writing a post on the topic from scratch (since engineering prompts can be its own rabbit-hole).
Big disclaimer: I am no expert in using ChatGPT or in AI (but there’s interesting advice about its use by other non-experts out there). Also, I am no futurologist and I have no idea how this technology will end up affecting the world of research (and of doing a PhD). Yet, I have the impression this will be a bit like the advent of the calculator: there will be upheaval for a time (especially, in the world of education and how learning is assessed), but eventually we humans will just shift our time and attention to creative activities that the new tools are not good at. For the time being, I think focusing on transparency and reproducibility of how we use these AI tools will probably be the key to an ethical use of these powerful new technologies. Thus, I have made the code for the algorithmically-generated parts of this blog post available in GitHub3.
A first try
Let’s see what the “default advice” of ChatGPT would be, which could represent what is in the “collective unconscious” of the Web-scale texts that went into the algorithm’s training. Using a simple prompt4 and temperature=0 (i.e., no extra chaos/creativity, the most reproducible response), this is what I got:
Boosting Productivity: Essential Advice for Doctoral Students
Introduction: Embarking on a doctoral journey is an exciting and challenging endeavor. As a doctoral student, managing your time effectively and maximizing productivity becomes crucial. To help you navigate this demanding phase, we have compiled some essential productivity advice tailored specifically for doctoral students. Let’s dive in!
- Set Clear Goals: Start by defining clear and specific goals for each day, week, and month. Break down your research project into smaller tasks and prioritize them based on their importance and deadlines. This approach will help you stay focused and motivated, ensuring steady progress towards your ultimate objective.
- Establish a Routine: Developing a consistent routine is key to maintaining productivity. Identify your most productive hours and allocate them to your most challenging tasks. Create a schedule that includes dedicated time for research, writing, reading, and self-care activities. Stick to this routine as much as possible to build discipline and enhance productivity.
- Manage Distractions: Doctoral students often face numerous distractions, both internal and external. Minimize interruptions by turning off notifications on your phone, closing unnecessary tabs on your computer, and finding a quiet workspace. Consider using productivity apps or browser extensions that block distracting websites during focused work sessions.
- Practice Effective Time Management: Time management is crucial for doctoral students. Break your work into manageable chunks and allocate specific time slots for each task. Use time-tracking techniques, such as the Pomodoro Technique, to work in focused bursts with short breaks in between. This method can help maintain concentration and prevent burnout.
- Prioritize Self-Care: Taking care of your physical and mental well-being is essential for maintaining productivity. Ensure you get enough sleep, exercise regularly, and eat nutritious meals. Make time for activities you enjoy outside of your research, such as hobbies or spending time with loved ones. Remember, a healthy mind and body are fundamental to academic success.
- Seek Support and Collaboration: Don’t hesitate to reach out for support when needed. Engage with your peers, advisors, and mentors to discuss challenges, seek guidance, and share ideas. Collaborating with others can provide fresh perspectives, enhance your research, and boost motivation. Attend conferences, workshops, and seminars to network with fellow researchers and expand your knowledge base.
- Embrace Effective Note-Taking and Organization: Develop a system for organizing your research materials, notes, and references. Utilize digital tools like reference management software (e.g., Zotero, Mendeley) to streamline your citation and bibliography process. Adopt effective note-taking techniques, such as the Cornell method or mind mapping, to capture and retain information efficiently.
- Emphasize Quality Over Quantity: While productivity is important, remember that quality research takes time. Avoid falling into the trap of focusing solely on output quantity. Instead, prioritize producing high-quality work that contributes meaningfully to your field. Take the time to thoroughly analyze data, critically evaluate sources, and refine your writing to ensure your research stands out.
Conclusion: As a doctoral student, productivity is crucial for successfully navigating the challenges of your research journey. By setting clear goals, establishing routines, managing distractions, practicing effective time management, prioritizing self-care, seeking support, embracing organization techniques, and emphasizing quality, you can enhance your productivity and make significant progress towards your academic goals. Remember, productivity is a journey, so be patient with yourself and celebrate each milestone along the way. Good luck!
Overall, pretty good, solid advice that I have given one time or another in the blog, and in our doctoral workshops. For instance, we have covered the bit about breaking down big projects into smaller tasks, and having daily and weekly goals/plans; we have talked about finding your most productive times; managing distractions (including addictions!); the Pomodoro technique was featured in one of the first blog posts I did (and still is a staple of my advice to doctoral students and to myself); we have covered sleep and other self-care tips; we have posts about how to get help from colleagues and how to structure writing collaboration; and I agree with the “quality over quantity” bit (see, e.g., this mantra).
Still, the text felt a bit generic, leaving me with a sense of having heard this all before, but no real clue of how to improve, or where to start. Or, maybe, too many clues (the list was a bit long for my taste). Probably doctoral students have received this sort of advice multiple times before, but if they are still reading about this topic maybe, for some reason, they haven’t managed to adopt such advice effectively.
Thus, as a (PhD student) reader, you can start with these 8 topics, see if you fall short in any of them, and read a bit more deeply on it (e.g., in the blog posts I link above, or other sources) and try to implement them better in your daily work. Also, I’d advice you to focus on one of them at a time, for a month or so, until you think you have developed good habits/systems around it.
As a writer and advice-giver, this not-so-bad text pushes me to find advice that is even more impactful, concrete and original. But, let’s give the algorithm another opportunity to push itself as well…
With some prompt engineering
After a few rounds of trying to make the prompt better (i.e., more concrete), I came to a much more detailed set of instructions for ChatGPT5. The result was:
The Surprising Productivity Principle for Doctoral Students
As doctoral students, we are often bombarded with advice on how to be more productive. From setting strict schedules to multitasking, the options seem endless. However, there is one contrarian principle that can truly revolutionize our productivity: less is more.
Less is More: The Key to Productivity
In a world where we are constantly juggling multiple responsibilities and tasks, it may seem counterintuitive to do less. But the truth is, by prioritizing and focusing on fewer tasks, we can actually achieve more. Here are three concrete pieces of advice to implement this principle in your daily life as a doctoral student.
1. Prioritize ruthlessly. Instead of trying to accomplish everything on your to-do list, take a step back and identify the most critical tasks. Focus on those that align with your long-term goals and have a significant impact on your research. For example:
- Humanities student: Instead of reading every single book on your topic, prioritize the seminal works and those that directly contribute to your research question.
- Science student: Instead of pursuing multiple experiments simultaneously, prioritize the ones that are most likely to yield groundbreaking results.
2. Embrace deep work. In a world filled with distractions, the ability to concentrate deeply on a single task is a rare but invaluable skill. Set aside dedicated blocks of time for deep work, where you eliminate all distractions and focus solely on your research. For instance:
- Social sciences student: Turn off notifications, close unnecessary tabs, and work in a quiet environment to dedicate uninterrupted time to data analysis or writing.
- Engineering student: Use time-blocking techniques to schedule uninterrupted periods for complex problem-solving or coding tasks.
3. Delegate and collaborate. As doctoral students, we often feel the need to do everything ourselves. However, learning to delegate and collaborate can significantly increase productivity. Identify tasks that can be outsourced or shared with others, allowing you to focus on your core research. Consider these examples:
- Medical student: Collaborate with colleagues to share the workload of literature reviews or data collection, ensuring a more comprehensive analysis.
- Business student: Delegate administrative tasks, such as organizing conferences or managing schedules, to capable assistants, freeing up time for strategic research.
By embracing the principle of less is more, prioritizing ruthlessly, embracing deep work, and delegating or collaborating effectively, doctoral students can achieve higher levels of productivity and make significant progress in their research. So, let’s break free from the overwhelming expectations and focus on what truly matters.
Wow, this is much better… even if it still sounds a bit generic and rehashing some of the ideas of the first attempt, like timeblocking and removing distractions. Especially the examples I found kinda lame and unrealistic (I wonder how many PhD students out there have “capable assistants” available… but if you do, by all means user them! :)). Again, I quite agree with the “contrarian principle” chosen, and have written about it, e.g., here and here… if anything, it does not really feel very contrarian, as the idea of “less is more” has been the target of a lot of recent productivity advice, in books like Oliver Burkeman’s “Four Thousand Weeks”, Greg McKeown’s “Essentialism”, or Cal Newport’s “Slow Productivity”.
Having a higher temperature (i.e., randomness, I used 0.7 in this final version) helped a bit with the “genericness”, but did not entirely remove this problem. Also, the highlighting that I asked for is not very good (beyond the structure I myself provided). It seems that the algorithm does not know very well which of the words it spits out is really more important than the others (which I guess should be expected).
Of course, I could continue engineering my prompt again and again, but then I would spend more time on that than I would have spent writing my own blog post. So I will stop here, and try to add some important “missing elements” from the algorithm’s output.
What the algorithm missed
Mainly, that “the devil is in the details”, as the old saying goes. How exactly to set clear goals? how exactly to know what to prioritize in our pursuit of doing “less but better”? how to communicate effectively when seeking feedback? it is not like we communicate badly or set shitty goals on purpose! Sometimes we do not know how to do things better. This is why, in many of the posts of this blog, we share concrete exercises or structures that have helped us and many other PhD students to do a variety of things, from writing the core of a paper idea or reflecting on the events of the past week to defining the key elements of our PhD and our doctoral journey.
Along the same line of the devil being in the details, there are specific time and task management tips especially tailored to doctoral students, which I found missing in ChatGPT’s advice. Maybe the most important of all is the importance of regularity and length of thesis-focused work sessions. Especially for part-time doctoral students, it is easy to end up peppering short (e.g., less than one hour) thesis-related bouts of activity across the week, or spend whole weeks without touching the thesis at all. Given the complexity of the ideas and tasks that a PhD often involves, this is a bad idea (by the time you finish loading the “thesis context” into your brain, the time is over). Having regular (e.g., multiple times per week) work sessions of at least 2-3 hours in length, are a good starting point. Another important tip is to limit your work-in-progress: not opening new threads of activity or projects until the current ones are closed (i.e., tackling big projects as sequentially as possible). A basic tenet of management strategies like kanban, basically this means focusing on just one thing (one study, one paper, one strand of literature reading) at a time until you finish it or can’t continue, before starting the next one. This will lead to less cognitive overload and stress (and it is often faster as well).
And finally, I would like to note one type of activity that is not mentioned often enough in doctoral advice: scheduling and spending time reflecting about both our PhD’s ideas and the projects/tasks that will make it a reality. This is not just about sitting on a couch with a glass of wine and thinking big, philosophical thoughts – these are structured reflections, like the ones happening at different time scales and frequencies to align our daily, weekly, and yearly plans with our goals at different time horizons, as well as periodic (monthly?) reflections about the key ideas in our thesis.
The diagram below summarizes the main ideas showcased in this blog post:
Over to you
Well, that was fun, at least for me. I hope it was interesting and enlightening for you as well…
Let me know if you want more ChatGPT-enhanced posts about other topics we cover in the blog, be it scientific writing or mental health. Also, do you have other particularly useful ways of using ChatGPT in your doctorate? Leave a voice message or write it in the comments section below!
PS: I’m experimenting with Amazon affiliate links, so if you click on the book links throughout the post and eventually buy the book there, I may get a small benefit (without any additional cost to you).
Header image by DALL-E via Bing6.
Prieto, Luis P., Paula Odriozola-González, María Jesús Rodríguez-Triana, Yannis Dimitriadis, and Tobias Ley. “Progress-Oriented Workshops for Doctoral Well-Being: Evidence From a Two-Country Design-Based Research.” International Journal of Doctoral Studies 17 (2022): 39–66. https://doi.org/10.28945/4898. ↩︎
Prieto, Luis P., Gerti Pishtari, Yannis Dimitriadis, María Jesús Rodríguez-Triana, Tobias Ley, and Paula Odriozola-González. “Single-Case Learning Analytics: Feasibility of a Human-Centered Analytics Approach to Support Doctoral Education.” JUCS - Journal of Universal Computer Science 29, no. 9 (September 28, 2023): 1033–68. https://doi.org/10.3897/jucs.94067. ↩︎
NB: I have tried to make this code as reproducible as possible (e.g., setting temperature=0 in some places – sadly, seed is not available in the R API I’m using). However, full reproducibility seems to be impossible, so your results WILL vary from what I wrote, probably. ↩︎
The initial prompt was: “Please write a blog post with the most important productivity advice you would give to a doctoral student. Please keep the text under 500 words." ↩︎
The improved prompt was: “Please write a short blog post with very concrete and surprising productivity advice for doctoral students. Please structure the text around a general but contrarian productivity principle, which is explained in the first part of the text, expanding it later to three concrete pieces of advice based on that principle, each one with a couple of realistic examples of how doctoral students from different disciplines can implement them in their daily lives. Write it in the voice of an English-proficient but not bilingual Spanish-speaking academic, making the tone collegial but correct. Take some time to think and structure your ideas before providing an answer. Please keep the text under 400 words, putting it out in Markdown syntax that highlights the main keywords and ideas." ↩︎
The prompt used was: “Please draw me a photorealist image of a doctoral student with eight normal and robotic arms, holding multiple lab instruments in a flurry of activity, wearing a labcoat. Use a 50mm lens and a wide aperture to blur the laboratory in the background. Use mostly dark colors." ↩︎
Luis P. Prieto
Luis P. is a Ramón y Cajal research fellow at the University of Valladolid (Spain), investigating learning technologies, especially learning analytics. He is also an avid learner about doctoral education and supervision, and he's the main author at the A Happy PhD blog.