Have you ever wondered how Facebook and Instagram algorithms affect what you see in your news feed?

To find out, Guardian Australia unleashed those algorithms on a completely blank smartphone linked to a new, unused email address.

Three months later, without any input, they were riddled with sexist and misogynistic content.

As invisible stakeholders drive invisible agendas which further widen societal and political divides, it should come as no surprise that the World Economic Forum Global Risks Report 2024 identifies misinformation and disinformation as the primary global risk in the short term.

Against this backdrop, as advertising and media strategists, we need to be hyper-attuned to our own unconscious bias. Both in terms of interpreting external data and insights and generating our own insights.

When it comes to generating insights, the impact of GenAI on advertising content and media strategy is the defining issue of the day. There seems to be no limit to GenAI ability to infiltrate and enhance creative and strategic expression.

But are there limits to AI cultural awareness and ability to generate culturally appropriate imaging?  Both audio and visual? Is GenAI basically just another tool for disseminating misinformation and disinformation?

Intellectual inertia

The local media industry’s intellectual inertia, when faced with the challenges of sub-optimal research data, prompted me to comment recently that strategists in Mzansi need to start thinking like The boy who harnessed the wind. We need to use what we’ve got until we’ve got what we want.

Now the real difference between me and Richard Kamkwamba is he had all the intellectual skills he needed to create electricity, but none of the tools. I’ve got all the tools I need to create AI content, but I lack the skills.

Despite this limitation, even I am beginning to enjoy the creative energy of GenAI.

Many of us in advertising have long upheld the adage “give me the freedom of a tight brief”. It’s a fundamental pillar of prompt engineering. So, providing the correct prompts for generating a picture of Audio Abe was really pretty easy.

Flushed with confidence after my Audio Abe triumph I prompted GenAI to create an image for me of a African boy building a “Heath Robinson” mechanism to interface with AI and the metaverse. In effect, based on the Richard Kamkwamba story, to create an image of the ‘boy who harnessed AI’. [See below left]

The response was immediate and I’m confident that seasoned strategists reading this would be able to discern the fundamental limitation of this image in the context of Africa.

So, I won’t labour the point.

One thing I do like about GenAI is that it can be very polite. I’m sure if prompted correctly AI would speak Gen Alpha.

So, you can imagine how impressed I was that instead of a “Hit me up if you want any tweaks!” I was asked “Let me know if you’d like any adjustments” to this image.

Enthused as I as by this exchange of civility I immediately responded

“Yes please. It’s Africa so could the boy more reflective of the demographics of Africa” …

“Thank you …”

The response from GenAI was immediate. [See image on right]

Again, I’m confident that seasoned advertising professionals will, at a glance, pick up the latent bias in the revised image. GenAI did change the image of ‘the boy who harnessed AI (or is that the bot who harnessed the wind?) but in doing so, they took away his smart clothes, his shoes and his mobile phone.

African reality or latent AI bias? You decide.

Maybe after the first image I would have a better chance of getting ratioed if I had just responded a Gen Alpha clapback like …