`` Nature '', a scientific journal with a history of 153 years, announces that it will ban the posting of images generated by AI

The scientific journal

Nature , founded in 1869, has announced that it will ban the publication of images generated using AI tools such as Stable Diffusion.

Why Nature will not allow the use of generative AI in images and video

Nature bans AI-generated art from its 153-year-old science journal | Ars Technica

On June 7, 2023, Nature published an article entitled 'Why we don't allow the use of generative AI in images and videos'. Nature notes that content created using generative AI tools such as ChatGPT and Midjourney has surged in recent years. As the functions of generative AI have rapidly improved, discussions, discussions, and consulting have continued within Nature for several months regarding the availability of generative AI.

And Nature said, ``With the exception of articles dedicated to AI, at least for the time being, Nature will not publish content that uses, in whole or in part, photos, videos, or illustrations generated using generative AI. For the time being, we will not use visual content generated by generative AI, with some exceptions.

In addition, artists, filmmakers, illustrators, and photographers who create visual content to be published in Nature must ensure that the work they submit is 'generated using AI' or 'created using AI.' It is required to confirm that it is not 'modified from'.

This is summarized on the following pages. Nature says, ``Exceptions are for images or art obtained from agencies with whom we have a contractual relationship that have created the images in a legally permissible manner.'' It includes images and videos that have been published, and will be reviewed on a case-by-case basis.' He also explained that there are cases where images created using generative AI tools can be posted.

Artificial Intelligence (AI) | Nature Portfolio

On why it would ban the use of generative AI-generated content, Nature said, 'At the end of the day, it's a question of integrity. The publishing process, as far as science and art are concerned, is underpinned by a shared commitment to integrity.' This also includes transparency: as researchers, editors and publishers, we need to know the sources of our data and images so we can verify their accuracy and truth. Existing generative AI tools do not provide access to the sources to perform such verification.'

In addition, attribution must be given when pre-existing work is used or cited. Nature pointed out that while this is a central principle in science and art, ``generative AI tools do not fit this principle.''

Also, if an individual is identifiable, or if the intellectual property of an artist or illustrator is involved, consent and permission must be obtained from the parties involved. However, in the case of general generative AI tools, Nature points out that this consent and permission process may also pose problems.

All of the above three points are problems that arise because `` generative AI is trained using images for which no effort has been made to identify the source ''.

Nature said, ``Works that are supposed to be copyrighted are routinely used for training generative AI without proper permission, and in some cases infringing on people's privacy.'' ``Privacy In addition to concerns about , the fact that deep fakes can be easily created also accelerates the spread of disinformation, ”he said, pointing out the problems of generative AI.

In addition, Nature does not allow AI to be the author of academic papers, saying that large-scale language models (LLM) such as ChatGPT ``do not meet the author criteria .'' However, we do allow the inclusion of text generated using generative AI in our papers with due diligence. However, this requires that the use of generative AI tools be documented in the methods or acknowledgments section of the paper, and authors must provide the source of all data, including AI-assisted generated data.

``The world is on the brink of an AI revolution. There are great expectations for this revolution, but AI, especially generative AI, is rapidly upending long-established conventions in fields such as science, art, and publishing. Yes, a revolutionary tool of this magnitude has taken centuries in some cases to develop, resulting in a system to protect the integrity of science and protect content creators from abuse. If we are not careful in handling AI, we run the risk of destroying all the achievements we have made so far.”

in Note, Posted by logu_ii