Though only four per cent of Canadian businesses say they are employing artificial intelligence, a majority of Canadian employees say they have used ChatGPT at work, according to a recent survey.
More than seven in 10 employees say they have used the tool at work, and more than eight in 10 said generative AI tools have increased their productivity, according to a survey of more than 600 full-time and part-time employees in Canada conducted by GetApp.
“Since a majority of users have found generative AI tools to be effective and innovative, this technology may drive even more traction in the coming years,” said Smriti Arya, an analyst for this study, in a press release.
“Employees in Canada find the technology more useful for text editing and text creation purposes as generative AI tools are usually trained on previously fed data to generate original content.”
Among the respondents who reported using ChatGPT at work, more than two in five (44 per cent) said they used the tool for text editing. Some also used ChatGPT for idea generation (37 per cent), data analysis (33 per cent) and composing emails (30 per cent).
But the assistance provided by ChatGPT and similar tools isn’t flawless. In some cases, the training data used by generative AI may be incorrect or out of date, which can result in wrong or duplicate content, according to Arya.
"It is ideal for organizations using generative AI to have some guidelines or strategies in place to ensure outcomes produced by generative AI are relevant and accurate,” he said.
The survey showed that nearly half (49 per cent) of Canadian employees using ChatGPT said they performed some checks on the generated outputs, although not for every response, while 42 per cent reported “meticulously reviewing and verifying every response before incorporating it into their work.”
Although ChatGPT was the most widely used generative AI tool among respondents, employees also reported using DeepMind's Alpha Code (22 per cent), DALL-E (15 per cent), Jasper (13 per cent), Midjourney (11 per cent) and Stable Diffusion (11 per cent).
As for what tasks generative AI does more efficiently, nearly half (46 per cent) of survey participants cited creativity. Around one third of respondents said generative AI can help with problem-solving, analyzing and summarizing data, and less than one third said they believed it can help perform repetitive tasks, research data and provide inspiration.