Editorial: AI versus natural stupidity

CHENNAI: Last week, international authors including John Grisham, Jodi Picoult, and George RR Martin joined a group of creators suing OpenAI for systematic theft of intellectual property (IP) on a mass scale. The episode is the latest in a wave of legal action initiated by writers concerned about artificial intelligence (AI) programmes using their copyrighted works without permission. In papers filed in a federal court in New York, the authors alleged flagrant infringements of plaintiffs’ registered copyrights, while referring to the ChatGPT programme as a massive commercial enterprise reliant on systematic theft on a gargantuan scale.

Martin, the author of A Song of Fire and Ice alleged that the programme generated an infringing, unauthorised, and detailed outline for a prequel to A Game of Thrones, that was titled A Dawn of Direwolves, and employed the same characters from Martin’s existing books.

Reacting to this, an OpenAI spokesperson said the company respects the rights of writers and believes they should benefit from AI. The company is now having conversations with the creative community to see where they go from here.

In August, OpenAI asked a federal judge in California to dismiss a lawsuit involving comedian Sarah Silverman, who called out the company, as well as Meta for mining through the contents of her memoir, The Bedwetter, without permission. The lawsuit was filed over the large language models (LLM) technology that underpins burgeoning generative-AI apps such as ChatGPT. As per the lawsuit, OpenAI and Meta have both programmed their respective LLMs partly on ‘shadow libraries’ — essentially repositories of vast amounts of pirated works of literature that are ‘flagrantly illegal’.

The onslaught of AI is far reaching in the arts world as seen lately. Rock legend Sir Paul McCartney in an interview spoke about how an AI-driven process had helped isolate the late crooner John Lennon’s vocals from an old tape. The star said that machine learning had helped lift Lennon’s voice from a demo and turn it into a song, which has now made its way into a final Beatles album. Apart from books and music, image-generation has also fallen under the scanner of AI.

Tools like Midjourney, DALL-E 2 and Stable Diffusion are now rewriting the laws of arts, photography and original ingenuity as we know it. Several magazines, newspapers and content generation portals, in an attempt to cut down on costs are deploying AI-based image generation tools to create realistic looking images and photographs, as an alternative to staging them in real studios or locations. What has happened simultaneously is that the business of stock images, and models/actors has taken a hit.

Corrective measures are being implemented as a result of a global outcry from the creative community. Amazon has instructed writers opting to publish through its Kindle Direct Program to notify the company if they plan on including AI-generated material. Microsoft has also announced that it will offer legal protection to customers being sued for copyright infringement over content generated by its AI tools. The US had also recently introduced a legislation in the Congress called the AI Disclosure Act of 2023. As per this, all material generated by AI tech would have to include a disclaimer regarding the same. And it would apply to videos, photos, text, audio, and/or any other AI generated material.