Since ChatGPT was released a year ago, people have not been able to shake off its craze. Released by OpenAI in November 2022, the chatbot, which generates incredibly realistic human-like text, appears to have permeated every industry, including science. Researchers make use of ChatGPT as well as what is called"Generative AI"A wider range of techniques to brainstorm ideas, come up with research ideas, create computers, or even write entire studies.
But not all scientists have embraced this technology. According to a survey conducted by the journal Nature, about 78% of researchers don't regularly use generative AI tools like ChatGPT. Many of the researchers who use it are simply used for recreational activities that are not related to research, or as experiments. Some people choose to steer clear of chatbots because of its potential pitfalls and limitations. Others fear they will miss out.
Nature interviewed three researchers about why they don't use ChatGPT at work.
If he could, Maxime Gauberti would definitely use ChatGPT. About a week ago, the neuroradiologist joined the waiting list for a subscription to ChatGPT-4. High Betty said"I tried to buy but it's not possible at the moment - I'm waiting for them to contact me. He had previously tried using an early free version of the AI tool, ChatGPT-35, but feel that its performance is not good enough, it is useless. "
One year after ChatGPT: who's using it, how it's being used, and why?
Gottty, who works at the Normandy University of Caon in France, believes that the latest version of ChatGPT can reduce the amount of time he spends writing emails and letters of recommendation for PhD students and postdocs in his lab. He said:"I think it would be really, really helpful for someone like me who is not a native English speaker. "I spent a lot of time writing to make sure that this was okay and that it was the right way to address someone. "
Speaking to other researchers who have access to ChatGPT-4, he emphasized how ChatGPT-4 has helped him. High Betty said"Now I see that some of my colleagues are using it, and I feel sorry because it takes so much time to do some tasks that can be automated. He often thinks about this, especially when he has to complete some tasks that could benefit from ChatGPT. "
Goberti believes that reducing the time spent writing emails can improve his overall productivity. He said:"It takes me at least two or three days to write a good enough email. If AI does this for me, I'll be more proactive, and if you multiply that by the number of times it might happen in your career, maybe it can have a little bit of a potential impact [on career development]. "
Stifling creativity?
Ada Kaluzna is a psychologist who left her research position at University College London in 2021 and now lives in Japan. She had tried to use ChatGPT to entertain her friends, but she didn't think the chatbot would be useful for her mental health research. "I love to write. I write fast. If I don't write my own research, then what am I going to be?
Carluzna believes that using ChatGPT could undermine her ability to learn and think creatively. She said"A lot of people say they don't know what they're thinking if they don't write their thoughts down. Writing allows you to focus on the topic and gives you time to brainstorm ideas. If you don't try to write down your thoughts in an organized way, you may not be able to come up with new ideas"。
While some of her colleagues use ChatGPT to write emails, Kaluzna says she's happy to avoid chatbots. She said"I don't feel stressed right now. "
Fear of fabricated information
Viswanath Vittaladevaram, a computational chemist at the University of Galway in Ireland, tried to use ChatGPT to find contextual information, but found that the information chatbots cite when answering niche questions is unreliable. He said:"A lot of the information generated through ChatGPT comes from non-academic**. "
He is also concerned when AI tools give different answers to the same question. Vitarad Valan said"I would feel like this is fabricated information. "He is also concerned that the data used to train ChatGPT will skew the chatbot's responses.
While Vitala Devallan tried to use ChatGPT to gather information, he didn't use it to write research** because he wanted to publish it in a scientific journal that banned the use of the tool**. Still, he believes that AI can be useful for answering basic questions such as the role of proteins in the human body. He said:"I decided to stay away from chatgpt altogether when it came to scientific writing. But I can use it to get some basic information"。
doi: