ChatGPT: five priorities for research

Abstract

A chatbot called ChatGPT can help to write text for essays, scientific abstracts and more. Since a chatbot called ChatGPT was released late last year, it has become apparent that this type of artificial intelligence (AI) technology will have huge implications on the way in which researchers work.ChatGPT is a large language model (LLM), a machine-learning system that autonomously learns from data and can produce sophisticated and seemingly intelligent writing after training on a massive data set of text. It is the latest in a series of such models released by OpenAI, an AI company in San Francisco, California, and by other firms. ChatGPT has

Transparency Guidance for ChatGPT Usage in Scientific Writing

“…This includes a better understanding of the capabilities and limitations of AI, as well as guidelines for its use and ethical considerations. By approaching AI with a balanced perspective, researchers can harness its potential while minimizing negative impacts 12 .…”

Risks and Benefits of Large Language Models for the Environment

“…This all could potentially accelerate scientific progress on important environmental topics. Conversely, these tools might prove to be a distraction, and researcher time may increasingly be consumed by dealing with adverse effects of LLM use, for example, in university education and training (e.g., misuse and cheating) . If LLMs are increasingly used to summarize information and to support funding decisions in the future, this could render human expertise and insight less important, potentially leading to adverse effects on research prioritization in the environmental sciences.…”

Why and how to embrace AI such as ChatGPT in your academic life

“…Like other powerful tools invented in history, such as the internet and personal computers, AI will have measurable short-term effects and potentially transformative long-term effects. As knowledge workers, it's in our best interest to embrace the power and potential of AI tools like ChatGPT to augment our skills, creativity, and productivity-ignore it at our peril 7 . Learning to write effective prompts to interact with AI is likely to become an indispensable skill; the prompts given in this comment (e.g.…”

“…In the context of research aid, a major limitation of LLMs is that they may fabricate facts, creating confident-sounding statements and legitimate-looking citations that are false ("hallucination"). Thus, as with any other source of information (e.g., Wikipedia), it is important to critically evaluate and verify AI responses, particularly when reliability is critical 7 . Additionally, although it is collaborative-meaning conversational-it tends to make assumptions about user intent based on the information given, rather than asking clarification questions.…”