More than a quarter of UK adults have used generative artificial intelligence such as chatbots, according to survey showing that 4 million people have also used it for work.
Generative AI, which refers to AI tools that produce convincing text or images in response to human prompts, has gripped the public imagination since the launch of ChatGPT in November.
The rate of adoption of the latest generation of AI systems exceeds that of voice-assisted speakers such as Amazon’s Alexa, according to accounting group Deloitte, which published the survey.
Deloitte said 26% of 16- to 75-year-olds have used a generative AI tool, representing about 13 million people, with one in 10 of those respondents using it at least once a day.
“It took five years for voice-assisted speakers to achieve the same adoption levels. It is incredibly rare for any emerging technology to achieve these levels of adoption and frequency of usage so rapidly,” said Paul Lee, a Deloitte partner.
The Deloitte survey of 4,150 UK adults found that just over half of the population had heard of generative AI, with around one in 10 respondents – the equivalent of approximately four million people – using it for work.
ChatGPT became a sensation due to its ability to generate human-seeming responses to a range of queries in different styles, producing articles, essays, jokes, poetry and job applications in response to text prompts.
It has been followed by Microsoft’s Bing chatbot, which is based on the same system as ChatGPT, Google’s Bard chatbot and, this week, Claude 2 from US firm Anthropic.
Image generators have also taken off, exemplified by a realistic-looking picture of Pope Francis in a puffer jacket, produced by US startup Midjourney.
However the ability of such systems to mass produce convincing text, image and even voice at scale has led to warnings that they could become tools for creating large-scale disinformation campaigns.
The Deloitte survey found that of those who had used generative AI, more than four out of 10 believe it always produces factually correct answers. One of the biggest flaws in generative AI systems so far is that they are prone to producing glaring factual errors.
“Generative AI technology is, however, still relatively nascent, with user interfaces, regulatory environment, legal status and accuracy still a work in progress,” said Lee.
