Ep 130: From AI Pilots to AI-Powered Banks - Ravi Bhat

Michael Avery (0:00)

South Africa has developed a really interesting habit when it comes to technology. We can be both much better than people imagine and not nearly good enough. Artificial intelligence appears just to be the latest example of that. According to Microsoft's latest global AI diffusion research, just over 23% of South Africa's working age population, so that's almost one in four, are now using generative AI. It's comfortably ahead of the global average of just under 18%, and it puts us ahead of every other African country measured, which sounds like real good news. And so you notice that South Africa still ranks only 46th out of 147 economies. And then there is an even more important question. What are we actually doing with that AI? We're using it to do what? Because asking ChatGPT to tally up an email or to summarize a document or plan dinner is technologically interesting. It's not necessarily an economic productivity revolution. The real prize arrives when companies redesign workflows, when banks process credit more intelligently, when developers build more software, and when fraud gets harder if you're in banking. All of those things are really what we should be looking for. It's a much higher bar.

I really like this as a use case, but it does raise a difficult question of what exactly happens to those 350,000 hours. And I think that's a good question. And I think that's a good question. And I think that's a good question. And I think that's a good question. And I think that's a good question.

Simply fewer people required to produce the existing output. I think that's a very important conversation for South Africa. But ultimately, there are lots of lessons we can learn. And joining me to unpack some of them is Ravi Bhat, who's Commercial Solutions and AI Officer at Microsoft South Africa. Ravi, welcome to The Monocle Banking Podcast. I've got to say, before we start, you know which episodes you're going to be watching?

The most recent episode on AI, just under 300,000. So it shows you there is a real thirst for knowledge around AI, but real use cases out there. I mean, how does that make you feel as Microsoft?

Ravi Bhat (2:44)

Michael, firstly, amazing to have been invited for your podcast. And yeah, you know what, I think as an AI officer, I am drinking from the fire.

I am drinking from the fire hose, just like everybody else in the world, because AI is progressing, evolving faster than many times humans are able to keep up. So I think it's an exciting world. We in Microsoft are creating a way forward for human race. We call it human-led AI, because we want to not be dino-sys and get extinct. We want to continue to be the controlling race, which controls AI going forward. So it's an exciting time for me, making sure that our engineering work, which we are doing, helps humanity. Just like our mission says, empower every person, every organization on the planet to do more.

Michael Avery (3:36)

That's a really interesting and intentional framing of human-led AI, because we have seen at a societal level, if you just look at the Pope's Magnificus Humanitas, if you look at the conversations around the way Frontier Labs might be hoovering up IP, there is a concern that humanity is maybe not at the center of this technology. But I don't want to get ahead of myself here. I want to start by interrogating Microsoft's own headline. We've reached this incredible AI adoption rate. So basically, call it one in four. Ahead of the global average, what exactly qualifies somebody as an AI adopter in that kind of number?

Ravi Bhat (4:14)

So AI adopter is somebody who uses AI in many lines of work in the organization, personally as well as at an enterprise level.

And the reason we are double-clicking on diffusion, the key point rather than just invention or creating chips as the point is because organizations in economies, countries who are diffusing and adopting AI faster, are seeing the benefits in the GDP in their AI economy. So when we talk about adoption or diffusion, it means actually use of AI in changing the way your work happens, your role happens, your organization.

Michael Avery (5:03)

Your role changes as your organization. Microsoft's broader research makes the point that AI leadership will increasingly be about depth rather than novelty. And I think that's quite clear, embedding these tools into everyday workflows rather than just producing impressive pilots. It was interesting when I was chatting to the CEO of 91, Hendrik de Toy, recently.

He still said that we're probably more in the impressive pilots phase and really embedding it just yet through the country's largest asset manager. What does depth look like inside an actual company from where you sit at Microsoft, Ravi?