As a researcher in academia, you know first-hand the challenges of managing your time and resources effectively. From conducting experiments and analysing data, to writing papers and grant proposals, there is a lot to juggle in a typical research career. Fortunately, chatbots like ChatGPT-3 (Generative Pre-trained Transformer) can be a valuable tool for helping researchers in academia streamline their workflows and get more done.
You might have seen viral posts about how ChatGPT-3 doesn't do so well, with its hilariously overconfidently wrong responses to some math questions and riddles. But, for the most part, in our experience, it has quickly become an invaluable tool.
But, firstly, what is ChatGPT?
ChatGPT is a variant of the GPT-3 language model that is specifically designed for conversational text. It has been trained on a large dataset of human-human conversations and is able to generate human-like responses to text inputs. Simply put, it is an AI-powered chatbot made by OpenAI that is free to use.
ChatGPT as your personal assistant and in-house copy writer
Think of ChatGPT as having a personal assistant or a copy writer. Have a bunch of emails you need to respond to? ChatGPT can help. By simply typing one or two sentences like 'Compose an email to person X replying to their question about Y with Z.' You'll be pleasantly surprised by the results.
A recent paper published in bioRxiv found that by just giving the title and journal of 10 papers published in high-impact journals, ChatGPT was able to generate believable abstracts that passed plagiarism tests. While the data it generated within the abstract was fabricated, you can start to see how amazing this technology is. A big timesink in academia is writing: writing emails, writing grant proposals, writing papers etc. While ChatGPT won't magically write these for you, it will at least give you a first draft, or a particular paragraph, and maybe help you over a writer's block. By using a chatbot to generate drafts of your papers, you can save time and focus on more important tasks, such as revising and editing your work.
After some time using ChatGPT, you'll notice the answers it generates sometimes feel formulaic and vague, but the foundation and first draft they provide allow the researcher to have a nice springboard to finishing a task they might have put off a bit longer.
ChatGPT as your academic pair-programmer
We've been testing ChatGPT and its ability to write code. This is perhaps one area where this type of technology excels. As academic research increasingly requires the use of analysis tools that necessitate a basic understanding of programming languages, it can seem daunting to add that skill to the list of things a researcher must learn or comprehend.
By typing a simple question or command into ChatGPT, like, 'write some python code to generate a graph for me' or 'write some code to generate a graphical user interface that increases a number with the click of a button' you'll quickly find that you have the foundations to a program that you can bulk out or use for these purposes.
It's also incredibly quick at giving you these responses. The above answer came back within seconds of submitting the query. It's easy to see its value and some are saying this could finally be Google's replacement.
Summarising information for you
Another amazing use-case for ChatGPT is its ability to concisely condense and summarise information fed into it. Have a science paper you don't have time to read? A particularly long email from a type-happy sender? ChatGPT can help you here, too.
Overall, chatbots like GPT-3 can be a valuable asset for researchers in academia, helping them to save time and focus on the most important aspects of their work. By automating repetitive tasks and providing access to relevant information and resources, chatbots can help researchers in academia work more efficiently and effectively.
Of all the brilliant advancements in machine learning and artificial intelligence applications of 2022, including Stable Diffusion among others, we don't think anything has captured the imagination of so many people quite like OpenAI's ChatGPT.