According to a report, Google has faced accusations of using OpenAI’s ChatGPT data without authorization to train its AI chatbot, Bard.
The report also notes that a Google AI researcher, Jacob Devlin, resigned after alleging that the company scraped ChatGPT data from ShareGPT, leading to his concerns about using such data in Bard’s training.
Devlin had voiced his concerns to Google’s CEO, Sundar Pichai, and other senior managers that the Bard team, with the aid of Brain employees, was using OpenAI’s ChatGPT data to train its machine learning model. Following his resignation, Devlin joined OpenAI to work on ChatGPT.
OpenAI and Google are direct competitors in the GenerativeAI field. Microsoft’s significant investment in OpenAI and its rapid integration of GPT into its products have pressured Google to launch its AI-powered chatbot, Bard, to the market. If the accusation that Google used ChatGPT data without permission is true, it could harm the company’s reputation, as it has invested considerable effort in AI research over the years.
Also, According to Loopex Digital, an SEO agency, Bard initially claimed to be based on OpenAI’s GPT-3 language model. However, in a subsequent exchange, Bard changed its statement and claimed to be based on Google AI’s LaMDA model. It’s possible this was simply a case of Bard providing incorrect information, a common flaw in generative AI models. However, if the opposite is true, it would lend credibility to the recent allegations. The agency shared screenshots of its chats with Bard for reference.
Google spokesperson Chris Pappas immediately responded to this news and has denied any association between Bard and ChatGPT data. Pappas stated that Bard “is not trained on any data from ShareGPT or ChatGPT,” thereby refuting the allegations against Google.
- Elon Musk, Yoshua Bengio, Steve Wozniak, Andrew Yang, And 1100+ More Signed An Open Letter Asking All AI Labs To Immediately Stop For At Least 6 Months