AI holds mirror up to gender and ethnic stereotypes in trades

Improving diversity is a key challenge for the construction sector.

Melanie Waters, managing director of trades body Trade-Up, said: “The fact that these AI tools favour depicting tradespeople as white men encapsulates the problem our industry faces as effectively as any statistic could.

“As AI is based upon gathering current and past content, it is never going to change unless the whole industry changes dramatically.

“The industry must now put boosting diversity among its workforce right at the top of its priority list, and that means asking difficult questions and looking outside the traditional recruitment methods.

“The Trade Skills Index report found that not only are we lacking women in leadership positions as are various industries, we’re lacking female representation altogether.”

Waters added: “These jobs are in huge demand in a stable industry offering superb security. They’re also rewarding, well-paid and come with great benefits.”

Checkatrade also plans imminently to launch its ‘Get In’ programme to encourage more young people into trades careers, including young women and those from ethnic minorities. 

Midjourney was asked to show ‘a builder’ and the results were exclusively white males. The app is known to produce different images from the same prompt and so it was repeated 40 times. Each time, the result was a white man. The prompt for electrician was similar, but out of the 40 requests it produced one woman.

When asked 10 times to ‘tell me a short story’ about specific trade jobs, ChatGPT related stories about men on nine out of 10 occasions. 

Don’t miss out on BIM and digital construction news: sign up to receive the BIMplus newsletter.