Chief AI Officer talks about course-correcting when AI is moving too fast

Machine Learning


In this episode, IT Leader Fast-5 — InformationWeek column for IT professionals looking for insights from their colleagues — Rob T. LeeChief AI Officer and Director of Research at the SANS Institute. that if the AI ​​moves too fast, it can “tilt everything”; Using a recent example from his own team. In Lee’s experience, rapidly implementing AI without organizational buy-in can create resistance and stall progress, as employees often “default to old processes.” We also explain why curiosity and a desire to keep learning are essential to working in today’s technology field.

Mr. Lee focuses on AI governance and security at the SANS Institute, a cybersecurity training, certification, and research organization, advising boards, CISOs, and government leaders on how to deploy and defend AI systems. He also serves as Forensic Technical Advisor to the Foreign Intelligence Surveillance Court through the U.S. Department of Justice.

Related:Use AI to select team leaders without crossing legal or ethical lines

important decision

What technical or organizational decisions have led to the biggest changes recently and why?

I’m really trying to make an impact not just in my organization, but in other organizations as well. learning concept. That’s not the only influence I’ve had [how to deal with] When it comes to cybersecurity, we’re trying to convince our friends, neighbors, and colleagues to admit how much they’re struggling. [with AI],that’s ok. Please go and try it. I have no intention of running a marathon tomorrow.

You should approach AI every day, just like you approach exercise. You should spend about 30 minutes every day learning the AI. It’s also about leading by example, which includes showing vulnerability by saying, “Hey, I’m having a hard time. ‘Don’t come to me with all the answers that I might not know.'”

Lessons learned the hard way

What recently didn’t go as planned, and what forced you to reconsider?

People use the old process by default. It becomes very difficult to become almost an “AI superpower.” The speed that AI can achieve.

Recent usage examples include the following discussion: [at SANS] About creating a microsite. We were told it could take anywhere from a few weeks to a month, and the team would need to do site design, AB testing, all of this.

I tried using ripple [AI] — and not being a trained web designer — I created a microsite in about an hour and a half. I shared it with my team lead to discuss it and they loved it, but then I shared it with the team that designs the website.

Everything tilted as old dogmas were challenged. At the same time, it demonstrated the potential speed of AI at a time when no one had yet caught up. There’s a palpable fear of how fast you’ve run and what that means for your team’s daily activities. There is a balance between the shock and awe of being able to take advantage of AI and the need for organizations to be mentally open to using these new technologies. I think this is something that most organizations will struggle with, even outside of cybersecurity.

Related:What Oracle’s headcount cuts reveal about running IT with fewer people

It causes disruption to the normal workflow… push/pull is definitely difficult. It’s not just fear of AI, it’s about jumping in the faces of people who said it would take weeks or months and saying, “Well, you’re wrong.” And people don’t like being told they’re wrong.

But I’m definitely sensitive to how it’s perceived by the current team. Implementing it and conveying the message correctly must be shown by the example of the leader. I may have failed because I didn’t properly communicate that to the other leaders. [The response] It was very polarizing, either very positive or very negative. Eventually, the organization said, “Let’s give it a try,” and everyone pitched in and worked on a solution.

talent trade-off

Related:How AI can enhance risk awareness and data insights for CIOs

Where are you currently investing in your talent, and what are you consciously not investing in?

It’s very easy to say that I want someone like this. AI expert But that’s too dualistic and doesn’t exist at the moment. What I’m really looking for is the ability and willingness to continuously learn, and the passion to become an expert at what I do.

A willingness to learn is almost the most important quality for a recruit. It is the ability to overcome the initial difficulties. It’s similar to exercising. Motivating yourself to participate in exercise can be difficult and can leave you feeling defeated. I always tell people that if you’re not frustrated, you’re not learning. If you’re not annoyed, you’re one of two different people. Either you already know that or you don’t care.

external signal

What recent external developments are most likely to change the way your organization operates, even indirectly?

Some recent developments [with AI] In reality, you’re trying to find workflows that can be optimized through the tools available to your employees. Collaborating with people to execute some workflows is essential, but it requires guidance. When I started dating everyone [on using AI]we use ChatGPT for basic analytics and also for creating and integrating email and Slack messages. AI could provide suggestions on how to improve the reception of your messages. One of the biggest benefits within our company is that we can change our cultural understanding of each other just by inputting basic communication through AI. So I start small.

change of perspective

Have you read, seen, or heard anything recently that changed the way you think about leadership or technology in any way?

I read this book through in about a weekend and listened to it on audio. This is a book by Pavlos Holman. Deep Future: Creating critical technologies.

It’s just a matter of getting into the nitty-gritty of how you need to look at different technologies and the concept of hacking to understand things. You can stop and look at things through a different lens.

My biggest thing is to never stop learning. Learning sounds very kitschy, but… [learning] It’s not just for work…it could be learning a new sport, painting, or playing the piano. Podcasts that provide inspiration and learning are also helpful to me.

Get the latest updates and insights from IT leaders straight to your inbox three times a week. InformationWeek Newsletter.





Source link