AI Optimism vs. Evidence in Late 2025 | Ylan Kazi (Blue Cross Blue Shield North Dakota)

**Laz** (0:08)
Data Storytellers, long-time recurring guest on the show, the one and only Ylan Kazi, and today we are going to take the opportunity to think through this moment in this whole enterprise AI shift. There's a lot of noise out there in the news cycle about what's happening, why it's happening, what might happen next, and, you know, we definitely don't have any kind of exclusive secret knowledge or access to that, but we are kind of plugged in into what's happening, especially on the front line of enterprises today, and it might help us to get closer to the truth. So, without further ado, Ylan, welcome back on the show.

**Ylan Kazi** (0:47)
Yeah, thanks for having me again, Laz. I'm looking forward to our discussion.

**Laz** (0:51)
Absolutely. So what we are going to talk about today, so there's a lot of, again, noise out there, and people are trying to figure out what might happen with AI. Okay, so AI is kind of dominating or has been dominating the new cycle for like two years now. It's a big trend of our age, or it seems so. And now a lot of people are talking about, oh, are we in an AI bubble? Was this overhyped? What will be the effect in the economy if it actually bursts and all that? So what I would like to do is, instead of just jumping in and kind of jumping the gun and trying to guess actually what might happen next, I would like to just turn back the clock and look at this whole thing from your perspective, because you were actually someone in the beginning who might have been perceived as an AI skeptic, right? When this whole hype started to grow. And I would like to understand that from your perspective, as an enterprise data and now AI leader, how have you processed this whole AI narrative? So if we just look at your role, right?
Your chief data officer at Blue Cross Blue Shield, North Dakota. When did AI become a significant narrative thread for you guys? And then maybe we can just kind of look at what happened before 2022, 2023 when Ched GPT revolutionized everything. Obviously, it wasn't Ched GPT, but for the average person, Ched GPT marks the point. So how did this whole AI conversation evolve from your perspective as a professional in your company?

**Ylan Kazi** (2:33)
Yeah, I mean, I started just over three years ago. So I think it was already being talked about, kind of in like broad brush strokes and specific use cases, that would really kind of help out our members and our stakeholders that we serve. But I think, like you said, once Ched GPT came out, that was really that light bulb moment.
And it wasn't even so much the functionality itself, it was the way that anyone with an internet connection could access it. I really think that was the key differentiator at the time. You know, if I rewind further back, 15 years ago, even 10 years ago, there was a lot more focus on how do you operationalize machine learning. And there was no real easy way of explaining that to stakeholders, and they couldn't just pick up their phone or their computer and start prompting it. Very different capabilities. And so at the time, there was still that excitement of, hey, if we apply machine learning or AI to something, will it make it better? And I think that's where my initial skepticism started, because I knew what it would take to actually operationalize it, test it, go through the entire end-to-end process. And it's so much more than pressing a button, right? A lot of people think, well, you just press a few buttons here or it's automated, it's good to go. I think that that skepticism has, it's helped me going forward. I think what's ironic, though, is I'm actually more excited about LLMs, the chat GPTs of the world, generative AI. In some cases, even more so than I was for traditional machine learning or traditional AI. So, I feel like there's just such a promise going forward, and we've really tapped maybe a fraction of that.

**Laz** (4:26)
Okay, so even before we go further, because we have listeners who are enterprise AI practitioners, people with a lot of technical knowledge or business savvy, but also we have listeners who maybe even today kind of struggle to see the distinction between quote unquote traditional AI. How traditional can you get when it comes to something like artificial intelligence? But what we coin or what we call traditional AI and generative AI, LLM. So, how would you articulate that in a simple way that might make us see the difference?