RM 196: Dr. James Maisiri on Responsible Technology Adoption

Mike Konrad (0:00)

Today, we're going to explore a topic that seems to be touching nearly every industry, every workplace, and increasingly, every part of our daily lives. What is that? Artificial intelligence. But this conversation is not just about what AI can do. It's about what AI assumes, what it values, what it misses, and what happens when powerful technologies are deployed into environments they were not designed to understand.

Welcome or hopefully welcome back to Reliability Matters. I'm so glad you're here. My guest today is Dr. James Maisiri, an AI researcher and writer whose work focuses on responsible AI, digital transformation, education, labor markets, and the societal impact of emerging technologies. His work has been featured through organizations and publications, including Unesco, Mail and Guardian, and many others. His message is both simple and profound. AI is not neutral. In one of his three TEDx talks, Dr. Maisiri said, if Africa does not shape AI, then AI will shape Africa. While his work often focuses on African communities and institutions, the lesson is much broader.

Every industry, including electronics manufacturing, needs to think carefully about how we adopt AI, what assumptions are built into these systems, and how we validate their use before trusting their output. In electronics reliability, we often say that context matters. Materials, environments, residues, humidity, temperature, field conditions and use cases all influence performance. Today, we'll ask whether the same is true for AI.

I'm not really sure how much of this conversation will tie in directly to reliability, but it's a fascinating subject nonetheless. So let's see where it takes us.

SPEAKER_2 (1:59)

Welcome to Reliability Matters, a podcast for the electronic assembly industry. Each episode covers topics related to reliability, best practices and environmentally responsible assembly techniques, with insights from experts across the electronic assembly industry. Now, here's your host, Mike Konrad.

Mike Konrad (2:18)

Hi James, welcome to the program. I'm excited to have this conversation with you today.

Dr. James Maisiri (2:24)

Hi Mike, how are you? Thank you very much for inviting me. I'm excited to be here.

Mike Konrad (2:29)

Yeah, as I'm excited that you're here as well.

Just so my audience knows where each of us are, I'm in California, you are not. Tell me where you are at this very moment.

Dr. James Maisiri (2:42)

All the way in Zimbabwe right now, on the other side of the world.

Mike Konrad (2:47)

The other side of there.

Dr. James Maisiri (2:48)

I'm in Zimbabwe in a city called, yeah, Bulawayo. I'm usually based in Joburg, but right now, I'm just in between places traveling, so I'm in Zimbabwe right now.

Mike Konrad (2:58)

There you go. You've said that if Africa, and this is a great title for a Ted Talk or a podcast. If Africa does not shape AI, then AI will shape Africa.

Tell me what that statement means to you. Where did that come from? What is the concern that that statement represents?

Dr. James Maisiri (3:27)

Yes. Right now, AI is the big craze that everyone is trying to adopt, even in African countries. In fact, in African countries, they believe that AI is going to help them leapfrog forward economically and catch up to the global north.

But my thinking was most AI systems globally are created in Western societies. China and America is 90 percent of the computing power globally that feed into AI.

So Africa just imports these technologies that are trained in Western societies. And we expect that the solutions that these technologies were created for that side, will have the exact same results within the African context. And as I was looking around, there were adverse consequences of just adopting AIs from the global north within the African context. And it was just a warning to the rest of Africa that we shouldn't be importing these technologies. We need to be creating our own AI systems with our own training of the African context to solve our problems. Because if we just import it here, we'll create more problems that we can even think of or to realize.

Mike Konrad (4:48)

Yeah, I think a lot of... That's an insight I hadn't really thought of before now.

Because AI can do many things. A lot of people use AI for math formulas. For example, if I ask AI, what is 2 plus 2, I'm going to get 4 And nowhere around the world would 4 be an inappropriate answer, right? When there's quantitative data that AI produces, it is probably accepted universally as a truth, no matter whether you're in South Africa or Ghana or wherever, or Mississippi or Canada or anywhere in Europe. People can agree that that's an accurate answer.

It's outside of quantitative data, I think, that you're referring to. And give me some examples, or an example, I guess, for the sake of time. I know you can probably come up with a thousand examples, of where Chat, GPT, or CLAWD, or any of the, you know, AI in general. I tend to say Chat GPT is just kind of the poster child. That's like Coca-Cola when I'm referring to Cola. But give me an example where AI might, based on Western training and Western norms, misapply some maybe advice to someone in Africa.