Despite buzz, generative AI tools such as ChatGPT see minimal daily use, survey finds

Despite buzz, generative AI tools such as ChatGPT see minimal daily use, survey finds

Since the launch of OpenAI's ChatGPT in November 2022, generative AI has been a hot topic in technology discussions, gaining widespread coverage for its capabilities.

But despite the hype surrounding the new artificial intelligence (AI) technologies, a recent survey reveals that few people regularly use AI tools like ChatGPT.

The survey, conducted by the Reuters Institute and Oxford University, included more than 12,000 individuals across six countries, including France, Denmark, and the UK, to measure attitudes towards using GenAI models.

The findings show that a minimal percentage of people use ChatGPT daily, with even fewer people using other generative AI tools such as Google Gemini and Microsoft Copilot.

Public familiarity with AI tools

A majority of respondents reported that they have heard of some popular AI tools.

ChatGPT emerged as the most recognised, with 61 per cent of respondents from Denmark and 58 per cent from the UK indicating they were aware of it.

However, similar tools like Google Gemini and Microsoft Copilot are less well known, with only about 15 to 25 per cent of people being aware of them.

The survey also found that more specialised AI tools, such as Midjourney and Perplexity, have little recognition among the general public.

People were also not more likely to be aware of AI products originating from their own country compared to internationally popular tools like ChatGPT.

For example, only three per cent of respondents from France reported being aware of Mistral, a French-based AI company.

While a significant majority of people have heard of highly publicised AI tools, fewer report actually using them, and regular usage remains rare, the survey found.

How people are using AI tools

ChatGPT is the most widely used AI tool among the people surveyed in six countries, being two or three times more prevalent than other large language models (LLMs) like Google’s Gemini and Microsoft Copilot.

However, the frequency of usage is not particularly high.

In France and the UK, only two per cent of respondents reported using ChatGPT daily, with even fewer using other AI tools; in fact, no one in the UK reported using Google Gemini daily.

Most respondents who have tried generative AI tools mentioned that they had only used them once or twice since their launch.

Meanwhile, the US shows higher usage rates for LLMs, with seven per cent of people using ChatGPT daily and 11 per cent weekly.

The study revealed that the most significant difference in AI tool usage was observed across different age groups.

Younger individuals were more likely to adopt AI tools and to use them more frequently.

In the six countries surveyed, 56 per cent of people aged 18 to 24 and 43 per cent of those aged 25 to 34 reported using generative AI at least once.

In contrast, only about 16 per cent of those aged 55 and over have used these tools.

Despite higher awareness and usage rates among younger populations, the overall frequency of usage remains low.

While most respondents reported using AI mainly in their personal lives rather than for work or academic purposes, the specific uses of AI varied.

Across all surveyed countries, 24 per cent of respondents used generative AI to gather information, while 28 per cent used it for media creation.

For those using it to obtain information, the most common uses include answering factual questions (11 per cent), seeking advice (10 per cent), and generating ideas (nine per cent).

Other uses for getting information include summarising text, support services, translations, and accessing the latest news.

On the other hand, those who use generative AI for media creation often employ it for tasks such as writing emails or letters (nine per cent), creating images (nine per cent), making videos (four per cent), and coding (five per cent).

Still, the majority of respondents who engage in media creation with generative AI report that they are primarily experimenting or playing around (11 per cent).

Anticipated impact of generative AI

Most people surveyed said they expect generative AI will significantly impact nearly every sector of society within the next five years.

According to the report, on average, 72 per cent of respondents anticipate that generative AI will have a major impact on search and social media companies.

Additionally, 66 per cent expect it will significantly affect the news media, and a similar percentage believe it will greatly influence the work of scientists.

Respondents also think AI will have a significant impact across various other industries, including healthcare and financial institutions (59 per cent each), the military (56 per cent), and governments (53 per cent).

Moreover, nearly half of those surveyed think these new technologies will have a large impact on ordinary people.

On average, most respondents expect generative AI to improve their lives to some extent, though a slightly higher number foresee a negative impact on society overall.