LeCun, a Turing Award-winning scientist, expressed support for using LLM to write papers. Veterans in the scientific research community may not like ChatGPT even more?

After the release of ChatGPT, whether "using language models to assist paper writing" is considered cheating and plagiarism, or a reasonable use of writing tools. Currently, there is still no unified standard among universities and paper conferences, and even different researchers have different opinions on ChatGPT. There are different attitudes.

Some people believe that ChatGPT can improve the quality of writing and allow non-native English speakers to reduce grammatical errors; however, some people believe that the generated text does not belong to the author and violates academic ethics.

Recently, Alexander Rush, an associate professor at Cornell University, was participating in ICLR 2023. During the conference exchange, he interviewed many participants about their attitudes towards "writing papers with LLM".

Judging from the results, the ratio of support and opposition is basically 50-50, and the degree of support and opposition is almost the same.

Should ICLR ban LLM from generating papers?

But if we divide it according to the groups of masters, doctoral students, and postdoctoral students, we can see that the number of years of education increases and the opposition rate is also rising. Among them, the group of masters students is more Tends to use it, but there are huge differences in the use of language models.

Researchers who have published more than 10 papers are most supportive of using language models to write papers, but the differences are also relatively large; in contrast, novices who have only published one paper.

Non-native English speakers and native English speakers generally support the use of language models to write papers, but native English speakers are even more divided.

Alexander Rush also asked during the survey "Should ICLR disable LLM to generate review comments?"

It can be seen that compared to using LLM as a tool when writing papers, more than 60% of researchers believe that LLM should not be used to review manuscripts.

And groups who have worked in the scientific research community for a longer period of time, such as PhDs, postdocs, and researchers who have published multiple papers, are also more inclined to ban LLM review.

Regarding the use of LLM to assist paper writing, we can see that the most common scenario is to correct grammar, but most people still only use LLM occasionally or intentionally to modify grammar.

In terms of related work research, Latex chart production, and writing summary of papers, the number of scientific researchers who use LLM as a daily tool has significantly decreased, but many people have expressed interest in exploring LLM in these fields. the use of.

Finally, the opinions of tutors and professors on banning LLM writing: Most people’s views are pros and cons, but generally speaking, the industry is more inclined to ban LLM, about 4% Professors believe that "restricting the use of LLM may be effective", and 10% of professors believe that "although it should be banned, some people still use it."

In short, Rush concluded after the survey that most people still have a wait-and-see attitude towards LLM as a scientific research tool, and the language model has triggered the Policy questions are important regarding issues related to "community trust," "cultural norms and biases," and "research and financial incentives."

Netizen evaluation

Turing Award winner Yann LeCun retweeted and said, As the co-founder and chairman of the ICLR Foundation, I am also a postdoctoral fellow and published There are more than 10 papers, but I'm a big fan of letting authors use the tools they find most useful to help them write technical papers , including large language models.

Most netizens are in favor of LLM being used in scientific research and believe that in the future, all scientific researchers must master the ability of AI collaboration.

Some netizens even joked that those who support the use of language models are fake researchers. Real researchers never use spelling check or calculators, but directly carve papers on stone. !

Some people think that ChatGPT is an artifact for writing Latex charts. In the past, it took several hours to find the answer on stack exchange, but now it only takes a few minutes!

ChatGPT is also very helpful in training scientific research skills. For example, if you use a calculator as a tutor, you can train your mental arithmetic skills.

And ChatGPT can discuss the topic of the paper with researchers, find loopholes, and improve the paper. This is a good thing. After all, not everyone has a genius partner around them to communicate with.

Do you think researchers should introduce LLM as a writing and review tool?

The above is the detailed content of LeCun, a Turing Award-winning scientist, expressed support for using LLM to write papers. Veterans in the scientific research community may not like ChatGPT even more?. For more information, please follow other related articles on the PHP Chinese website!