Diversity in tech: To work for all, AI needs input from all

If there’s one sector of the business world that has benefited most from the emerging artificial intelligence (AI) revolution, it’s market research. Rarely have there been so many surveys commissioned by tech suppliers attempting to capture this moment  – often, strangely, in ways that happen to coincide with their product marketing. Weird, right?

According to a small selection of recent studies pitched to Computer Weekly, 58% of UK business decision-makers do not currently view AI adoption as a strategic priority  – although 58% of UK businesses are now, apparently, introducing generative AI (GenAI). Meanwhile, 59% of UK business owners believe adopting AI could help improve efficiencies, but 46% admit they are overwhelmed by AI and unsure where to start.

Even the roguest of the many rogue GenAI hallucinations might struggle to produce such consistently inconsistent conclusions. Nonetheless, the burgeoning AI industry has the ear of governments and boardrooms the world over, convincing them that AI is the solution to global economic struggles.

This is one reason why this year we chose to theme our annual diversity in tech conference around the impact of AI on this vital but perpetually under-invested area. In a room full of leading – largely but not exclusively female – tech leaders, there was unanimity that the progress of AI is being made at the expense of under-represented voices and talents. With such uncertainty and confusion in the market – a market being led by the vested interest of AI suppliers – perhaps that’s hardly surprising.

Maybe, just maybe, one of the reasons AI is confusing buyers and users is because it does not reflect the world as they experience it? The biases and assumptions built into many – all? – AI models comes directly from the lack of diversity among those developing, building and selling it.

There are huge risks from hurtling headlong into the AI era without assuring it will reflect the full diversity of the world it hopes to change. The UK government is predicating billions of pounds of savings in public spending on promises from AI companies over productivity improvements – yet the government also seems reluctant to listen to tech practitioners with coalface experience of implementing new technologies and the challenges involved. Imagine the Post Office scandal, multiplied by AI in lacking the understanding of the very real, very diverse people whose lives will be directly affected by it.

There are already many reasons why the tech community needs, finally, to address its diversity problem – a need the industry sadly appears to be shrinking from. But if AI is set to accelerate business, economic and social change, it needs first to accelerate diversity and inclusion across the industry that is so eager to profit from it.