Tech giant Google denies copying OpenAI’s ChatGPT to train its AI chatbot called Bard AI
Google has refuted claims that it used OpenAI’s ChatGPT, which is owned by Microsoft, to train Bard, its AI chatbot.
According to a report in The Information, the two AI research teams at Google’s parent company, Alphabet, had been fiercely competing for years until being driven by OpenAI’s success in collaborating.
The article cites sources as saying that personnel at DeepMind, an Alphabet subsidiary firm, are collaborating with software developers at Google’s Brain AI branch to create software that will compete with OpenAI.
The combined project, referred to as Gemini internally, started recently following Google’s failure with Bard AI, its initial attempt to take on OpenAI’s chatbot, according to the article.
The Verge was informed by a Google representative that “Bard is not trained on any data from ShareGPT or ChatGPT.”
In the meantime, Google has said that it will allow users to interact with its ChatGPT rival “Bard” as part of an early experiment.
The US and the UK already have early access to Bard, and the business promised to add more nations and languages as time goes on.
Bard is built on a large language model (LLM), especially a lightweight and optimized version of LaMDA, which the tech giant stated will be upgraded with newer, more competent models in the future. It is similar to OpenAI’s ChatGPT and Microsoft’s Bing chatbot.
Users can communicate with Bard by posing queries and then clarifying their answers with further queries.
The post Google Denies Copying OpenAI’s ChatGPT to Train its Bard AI appeared first on Analytics Insight.
