According to a report from The Information, Google allegedly uses data from OpenAI ChatGPT without permission to train its AI chatbot, Bard. The report claims that Jacob Devlin, a Google AI researcher, resigned after raising concerns with senior managers that the Bard team, backed by Brain employees, was using ChatGPT data to train machine learning models. Devlin has since joined OpenAI and is reportedly working on ChatGPT.
OpenAI and Google are in direct competition in the field of generative AI, and accusations of Google’s use of ChatGPT data could damage the company’s reputation. The Information also reports that Google’s Brain AI team is working with another of Alphabet’s AI companies, DeepMind, on a new project codenamed ‘Gemini’ to develop a product to compete with OpenAI’s efforts.
Google spokesperson Chris Pappas denies that Bard is based on ChatGPT data, stating, “Bard does not train on data from ShareGPT or ChatGPT,” to The Verge.
The report raises concerns about the ethical use of AI models and the importance of obtaining permission for data usage. Using unauthorized data from a competitor could lead to legal action and harm the relationship between AI research organizations. The report also highlights the intense competition among major technology companies to develop AI models that produce human-like responses and how this competition could lead to questionable practices.
As AI continues to grow and evolve, companies must prioritize ethical practices and transparency in their AI development efforts. This includes obtaining proper permission for data usage and avoiding practices that could harm the company’s reputation or the whole AI industry.