Google has revealed a major revamp to its artificial intelligence (AI) chatbot as it seeks to rival OpenAI’s ChatGPT.
The US tech giant’s Google Deepmind division has unveiled Gemini, an upgrade to the Bard chatbot it launched earlier this year.
Sundar Pichai, chief executive of Google’s parent company Alphabet, said the new AI technology represented “one of the biggest science and engineering efforts we’ve undertaken as a company”.
He said its Gemini upgrade will add more coding and mathematical abilities to its existing Bard chatbot, while it will also be able to recognise and describe images, video and audio. A streamlined version of its AI will also work directly on mobile phones, without the need for an internet connection.
Demis Hassabis, the chief executive of Google Deepmind, said the most advanced features would initially be available for a handful of customers for testing, before being expanded widely early next year.
It is understood UK government officials will be given advanced access to Google’s most powerful AI models as part of work for the AI Safety Institute.
Sissie Hsiao, general manager of Google Bard, said: “We’re currently completing extensive safety checks and will launch a trusted tester program soon before opening Bard Advanced up to more people early next year.”
Google executives declined to disclose how powerful the new chatbot was but claimed it outperformed the free version of ChatGPT launched in November 2022.
Google has been racing to compete with ChatGPT, which became the fastest internet service ever to reach 100 million users this year.
It is currently attracting north of one billion users each month to its web and smartphone apps.
The sudden success of ChatGPT prompted a “code red” inside Google as executives scrambled to catch up with the smaller rival.
The upgraded version of Bard will be available in 170 countries, including the UK and Europe at a future date.