Print This PostArtificial intelligence, chat bots, and social media: integrating each new wave of technology into our everyday lives often seems both overwhelming and somewhat unavoidable. So, even if ChatGPT itself will admit to the wide range of practical and ethical concerns at play in this increasingly complex world of tech, I thought it would be useful to take away some of the mystery and focus instead on pragmatic ways we can implement these new tools into our deep-routed routines.
First up, ChatGPT itself. Large-language models (LLMs) are an increasingly accessible tool which are able to scour the internet for information, often at much faster rates than traditional search engines. Are they always correct? No! If you were trained from birth on the best and worst of the World Wide Web, I dare say that you’d have some quirks as well. But like the internet before them, LLMs are a useful tool in collating information quickly and finding jumping-off points for future research. Perhaps more excitingly, this also extends to code busting. Instead of scrolling through pages of Mathworks in the hope of finding a related problem to the error message haunting your nightmares, why not venture a question to your favourite LLM? I’ve personally had instances where ChatGPT will fabricate fictional functions, but I’ve also had it recommend real packages which I didn’t know existed. I often describe LLMs as “the internet on steroids” and thus I treat them with the same level of enthusiasm and caution. Everything might be taken with a grain of salt, but at least I have recipe suggestions to get me started.
With that said, if you’re still a little wary of the decidedly black boxes of ChatGPT and co – why not switch over to a slightly more academic approach? Sites like Elicit.com, Consensus and SciSpace allow you to ask scientific questions and receive a synthesised and referenced answer based on available literature (see below). These are made much more with scientists in mind and each has its own unique functionalities. For example, Elicit.com will tell you how strong your research question is, and provide suggestions to refine it further. Consensus adds badges to identify different types of research and highlights articles which are highly-cited or from rigorous journals. What they have in common is that once you’ve asked your question you can scroll through the recommended list of literature yourself, with all of your usual scientific cynicism. As with sites like ChatPDF, Consensus and SciSpace also include a feature that allows you to upload your own papers and ask them questions directly. All useful techniques in finding and synthesising literature quickly without sacrificing on knowing exactly where it comes from.
Consensus on ChatGPT in academia: Most studies report mixed views, highlighting both potential benefits and ethical concerns.
Once you have a few papers to serve as starting points, you can also head over to sites like Litmaps or Connected Papers. Type in a paper DOI and these sites will provide a network of papers connected to your original, organised by the strength of the connection, date of publication and number of citations (see below). Especially when trying to get the lay of the land for a new topic, these maps can be incredibly effective in ensuring no particularly impactful paper is left behind. All of these sites also have privacy policies you can review to check what information they collect from you and your PC.
Citation map showing key academic papers and connections on ChatGPT’s role in education.
A final note on the topic of reading that always bears repeating – if you aren’t using a reference manager, please reconsider your life choices. Zotero is free to download and comes with extensions to Word and your favourite search engine meaning you can add papers to your digital library with the click of a button and reference them in a wide range of styles without typing out every word of the bibliography by hand. As someone who manually referenced until her PhD days – learn from my mistakes!
If you take nothing else away from this blog, please download Zotero.
So, you’ve researched your article and stored all of those wonderful papers into a reference manager ready for the writing process to begin. There are a number of AI tools which can also be helpful for the writing and editing stage, such as Paperpal and JenniAI, but for me this is the boundary of implementing AI in my academic life. I’d always rather sit and write with the power of my own little head and the literature I’ve accrued. With that in mind, I can’t recommend AI writing tools with the same enthusiasm as reading and referencing tools, though there are plenty of other resources available on this topic. When it comes to generating figures, I stand by PowerPoint as my favourite graphics design tool of choice, with some software (aka JASP) even allowing you to save figures directly as PowerPoint files ready for editing. If you haven’t explored the icons and illustrations freely available with more recent versions of PowerPoint, I can highly recommend them. It’s amazing what you can do with just some shapes and a vision. But if you want something slightly fancier, there’s always the BioRender and FlatIcons of the world ready to help out.
Finally, once you have your paper published – it’s time to share it with the world. Social media is a large player in the tech industry which can be helpful in increasing the reach of your work, alongside collaborating with your institution’s communication team. Though there are long established platforms like X still available, the new kid on the block is BlueSky. If you’ve been waiting for the perfect time to join this newest offering on the social media market – consider this your sign to hop aboard! A few things to know before you get started:
- BlueSky operates on hashtags in a similar way to many other platforms – the first ones you’ll likely want to add are #neuroskyence and #AcademicSky. From there why not try #CogSci and #WomeninSTEM, as well as following Dementia Researcher of course.
- Starter packs are your friend – many researchers have put together packs of BlueSky users who focus on particular subjects, such as dementia. This means you can follow tens or hundreds of relevant researchers with a single click rather than adding everyone individually. Just use the usual search function to seek out those most relevant to you or check out this list from The Transmitter.
- Get engaging! As with most social media, you get back what you put into BlueSky. Especially at the beginning it may feel a little like a ghost town, but once you start curating your feed with hashtags and following people from your academic community, hopefully #neuroskyence is something we can all grow together.
With that, the robot overlords are telling me it’s time to wrap up. The truth is that even as a young researcher, I’ve seen the technology sector rapidly shape the way I engage with academia on a day-to-day basis, making me more efficient and perhaps even a better researcher.
In the words of ChatGPT: “The key is to embrace the good while staying mindful of the risks”.
As with every wave of new technology, it takes a while to find our individual balance and make these modern tools work for us. But I hope this blog may have at least given you a few new ideas of how to leverage the tech of tomorrow, to improve your everyday.
Rebecca Williams
Author
Rebecca Williams is a PhD student at the University of Cambridge. Though originally from ‘up North’ in a small town called Leigh, she did her undergraduate and masters at the University of Oxford before defecting to Cambridge for her doctorate researching Frontotemporal dementia and Apathy. She now spends her days collecting data from wonderful volunteers, and coding. Outside work, she plays board games, and is very crafty.
Dementia Researcher
Dr Emma Law
Dr Clíona Farrell
Rebecca Williams