License to Trade in the Informational Age: Idealabs' Attribution Algorithm, Protection Racket or Creator Pushback
The Forefathers of today's Al in the olden days of computing (1950s) understood Al to be a simple transaction of an” if this - then that” statement, a system that learns upon the previous transaction and then builds upon its own recognition abilities according to its designed topic recognition. The result is a system that runs infinite computations of comparisons until a model emerges with a steady high rate of detection.
Attribution
Bill Gross, chair member of IdeaLabs, created ProRata, a chatbot and search engine trained to detect and attribute sources of AI models. Media giants like Universal Music, The Atlantic, The Financial Times, and Axel Springer have pledged support to the site and licensed their own content to the startup’s attribution algorithm as Mr. Gross and his partners attempt to crack down on open-source generative AI’s learning methods. ProRata aims to detect sources of generated content back to their original content creator by deconstructing the elements of generated content with an attribution measuring algorithm, “taking the output of generative AI, whether it’s text or an image, music or a movie, breaking it down into the components to figure out where they came from, and then providing a percentage attribution to each copy title holder.” The algorithm measures the value of the contributing content and calculates a proportional value, which ProRata claims is crucial for revenue to be apportioned correctly to the rightful copyright holder “even when multiple sources are used in a generated answer.”
In the face of increasingly rapid advances in the abilities of generative AI, ProRata capitalizes upon the fear of destruction for the creationist spirit and their representing groups’ desire for defense. Naturally, the content creators’ fear of having their work stolen or replaced by their own tools and the representatives’ loss of profit characterize this trend. ProRata claims to that every creator gets their “fair share of the pie.” Its goal is to circumvent the endless legal battles between media groups, creators, and AI companies accused of stealing data from publicly reposted content to their generative models. Creators would have to levy endlessly against AI companies, whose practices are alleged to be stealing data from websites, videos, and captions reposted publicly on the internet for the purposes of generative bot training amounts to theft, shoplifting or plagiarism.
Mr. Gross, the inventor of the original pay-per-click advertising system, aims to now set a new revenue-sharing system between creators, their publishers, and generative AI companies in a relationship he expressly claims is “not a license to trade” in the information age. However, Mr. Gross’s business practices and statements say otherwise. Mr. Gross believes that if he can get enough people to sign up for ProRata’s all-licensed-data AI search engine, he can make enough money to pay data providers their allotted share.”
Companies and longtime media executives are quick to support Mr. Gross’ new method to curb the free use of copyrighted material by AI for generative production. However, Mr. Gross’s approach creates a preemptive protectionist racket, fueled by publishers’ fear of profit loss. While Mr. Gross’s motivations feel more egocentrically grounded in the pursuit of lining his pockets, his algorithm may be the solution copyright law needs to circumvent the drawn-out battles fought between individuals and globalized forces: make everyone subscribe to one system.
However, the Atlantic already signed its own deal with OpenAI in May of 2024 to make generated information “discoverable with links and attribution within ChatGPT.” The Atlantic’s deal was an attempt to bridge the gap between properly crediting and compensating the creators and publishers of the work AI relies upon. Yet due to executive and employee backlash, authors for The Atlantic wrote and titled a letter to Chief Executive Nicholas Thompson and Editor-In-Chief Jeffery Goldberg on August 1, 2024, stating that “The Atlantic is a magazine made by humans, for humans.” Since August, The Atlantic has switched its stance on its deal in May with OpenAI, partnering instead with ProRata’s content control, permission, and clear attribution approach.
Other media bodies, such as Fortune, have already partnered with different AI attribution and revenue-sharing plans. Perplexity, an AI chatbot similar to ProRata’s, is the first of many new AI bots used to track and attribute creator and publisher copyrighted content. It appears to signal a new generation’s industrial norm and practice for information that the law will have no choice but to catch up to.