Generative AI in the Spotlight: Innovations, Biases, And Ethical Challenges

Did you see how Generative AI has transformed our everyday selfies into Renaissance-style portraits or transport us into the future by envisaging our prematurely aged faces? Generative AI is burgeoning, making big leaps forward in natural language processing where it masters the art of writing poems and churning out captivating content that bounces off from just about any theme.

Nevertheless, what lurks beneath these mind-blowing breakthroughs are inherent risks we need to be wary of. With such opportunities Generative AI has given us, comes the power it wields to those who manipulate. As it is just an algorithm that is pretrained with large amounts of datasets, this can unintentionally lead to the rise of skewed and biased content.

We interviewed Mal Chia, Managing Director of Ecom Nation, to find out more about his opinion piece on the political and ethical issues when Generative AI has been woven into the fabric of businesses, the rise of job displacement when hoping to implement it at scale, as well as enhancing data privacy and security.

What are some of the ethical implications or unintended consequences when implementing Generative AI and how can we mitigate them?

I recently stumbled upon a fascinating New York Times article that tested a bunch of common Generative AI platforms for political bias. It turns out these AIs have distinct political leanings, all shaped by the data they’ve been trained on. As they're large language models pulling from a vast array of sources, they naturally start to show a bias. It's vital we understand this bias. For example, if we ask MidJourney or ChatGPT to generate an image of a CEO, odds are you'll get a middle-aged white male. That’s the bias right there. We need to be savvy with our prompts, ensure we're providing the right cues to steer the AI and counteract these biases as much as we can.

What are your opinions toward the threat of Generative AI as some claim that it can displace our skills due to its nature of being automated?

This whole debate about AI taking over our jobs? It’s not new. Back in the 1990s with the dawn of the internet, or even the early 1900s during the Industrial Revolution—people were freaking out then, too that technology was going to take away jobs. What happened? Jobs evolved. We are in a period of historically low unemployment. Instead of assembling cars by hand, we started operating the machines that do. Humans are incredibly adaptable. I’m optimistic we’ll just keep evolving. So, I don’t see AI as a threat to our ethics or jobs per se, but more as a push for us to level up and adapt.

When AI analyses large troves of data, data privacy and security are some areas retailers need to be aware when they integrate it in their operations. What do you think of this?

Integrating AI into operations brings up big questions around data privacy and security. Key questions being: Where is the data going? Where is it stored? I'm particularly cautious about ensuring that any data I upload is anonymized—I don’t want to accidentally leak sensitive info into an AI model. There’s also the issue with deepfakes—already being used to create misleading images of celebrities or politicians. It’s critical that we hold platforms accountable for labeling AI-generated content accurately. This is manageable in the U.S., but what about when the content's coming from nefarious actors from Russia? They're not playing by the same rules. We've got to enforce these standards wherever we can, keep things in check and help people to spot the difference.

Hear from Mal Chia at Equarius Hotel, Singapore, on 15th May, 12:45PM: Keynote Interview: Separating the hype from tangible advancements – Join this session to delve into the reality of Generative AI, moving beyond the buzzwords and addressing the genuine impact it holds. Find out more here!