Noelle Russell explains why leaders need to move beyond AI enthusiasm to responsible execution. Her book's tiger symbolizes the eager early phase of AI adoption that must mature into thoughtful implementation. Russell argues that mastering responsible AI practices ensures technology serves humanity safely and ethically.
I do think it's about making AI work for us responsibly, which is the theme, the crux of your book. And I love the subtitle, From Enthusiasm to Execution. Tell us a little bit about that.
Yes, absolutely. And you'll notice on the COVID of the book is a tiger. I was very proud that they put the tiger on the book. So cute.
I'm holding up my book for those listening.
I have this analogy that actually came to me when I was working at Microsoft. My role was to really help researchers take their products and put them into production. And what I saw in those early days of taking products and giving them to users is exactly what we ended up seeing. And this is like in 2018, 2019. But in 2022, we saw the same thing happen as ChatGPT unleashed into the world. People got excited. They were like, I always encourage people to just think about the first time you used a large language model. You're like, well, that's pretty good. I can't believe it did that. That's pretty awesome. And I call that kind of baby tiger mode or like the peak of enthusiasm. Like you're like, this is amazing. Everything's awesome. But very few people are actually starting to see this happen now at scale is like very few people in this stage of like baby tiger mode. They're not really asking important questions like, Hey, you know, how big is this thing gonna get and where is it gonna live and how, what does it eat and how much data is it gonna need and where, you know, what happens if I don't want it anymore, right? We can think about many organizations that are failing and they have massive amounts of critical personal data. What happens to that data? Who owns that now? Can I just buy it on the open market if I'm a rich billionaire looking for DNA data? Like it's a very concerning, time. But if we ask those questions at the beginning, we can create mitigations, we can create security. I always say we want to weave it into the DNA of a project. That's why agentic is another term, you know, not to take people down a rabbit hole, but this concept of agentic is actually a good idea because it creates an opportunity for us to build little AI models that can do the work of responsibility, that can build responsibility into an application so that when you actually gets sold or acquired, or the team gets laid off, those systems will have it built in. You can't pluck it out like a raisin in a bun, right? Like, it'll be like water in a wave. And I think that is what this whole concept of kind of baby tigerness came from. This concept of enthusiasm and execution is that in the moment that you're excited is the most important time to ask, what is it going to take to actually execute at scale? And those questions are often, not technical ones. They're mostly organizational ones, right? Who's gonna lead this? Who's gonna decide what's going and what isn't? And ultimately, who's gonna tell us, you know, to stop if it's not a model that serves everyone?
