AI’s biggest challenge isn’t the technology

If you've been following my posts, you will know I've spent a fair amount of time doing AI coding over the last six months. More than anything, the countless hours I spent working with AI gave me a deep understanding of just how different it is as a technology. As much as I had to learn, there were also some things I had to unlearn.

And as organisations turn to AI to compete effectively, employees are placed in the unenviable position of adapting a lifetime of learnings to the era of AI. What mindset should they adopt to forge ahead instead of floundering? And how can organisations help them?

In the second episode of the Why Memory and Storage Matter podcast with Micron Technology, we explore the human side of AI adoption, in other words, where AI is actually built.

Change is already here

While I typically focus on the technical side of things, I really enjoyed this conversation with Cher Whee and Dane Tang. And the dynamics worked great: Cher Whee is the Vice President, People Strategy, Technology and Workforce Transformation at Micron Technology, while Dane is currently a leadership coach with two decades of experience at large MNCs.

Dane shared an anecdote that will probably resonate with more than one of you. A senior leader she works with recently texted her about how a junior engineer using AI did a task in 10 minutes that used to take him two months. This story highlights how dramatic AI’s gains can be. It also raises the question: how do we add value? Are we still relevant?

Cher Whee, who started her career as an engineer, offers a way forward. In her view, every job will change, and the opportunity lies in what we make of it: “All jobs will eventually evolve and transform. That ability to bring meaning to the job is where we are going to unlock the potential of humankind.”

Part of that, she says, is recognising that AI fluency isn’t about everyone learning the same skills: “We will never attempt to make everyone a data scientist or a prompt engineer. What is important is that the tools are meaningful to the role they are playing – and knowing when to trust the AI output, and when not to.”

Building fluency, role by role

According to Cher Whee, research shows that just 30% of an AI transformation is the technology. The other 70% is people. That’s why Micron meets employees where they are. Anyone who asks for an AI tool gets one, and a fluency assessment tells each person where they stand, so training can start from there.

But what does AI fluency look like in practice? That depends on who you are. For an engineer, Cher Whee explained, it’s about accelerating problem solving. For an operator, it’s using the tools at hand to report gaps against the specs they’ve been trained on. And for leaders, “what really matters is the quality of conversation and coaching.”

Dane likens AI fluency to being bilingual: “The first level is your core competency: HR, finance, marketing – be very good at what you do. The second is AI, to understand the tools and use them to amplify what you do.”

I wholeheartedly agree with Dane here. In fact, it’s why I took up AI coding in the first place. I wanted to augment my existing experience and skills in storytelling and editorial strategy with AI. What capabilities can I build that AI alone can’t match? And how can I use AI to do far more than the old me ever could?

Measuring what matters

But handing out tools isn’t enough. Cher Whee was candid that building an AI solution on top of a poor process won’t deliver impact. To tackle this, Micron goes back to the foundation and redesigns the workflow first.

It starts with a set of difficult questions: “Is this step even needed? Can we stop doing certain parts? Can we simplify? Can we automate before we build an AI agent?”

Measuring success calls for the same rigour. Counting licences and logins isn't effective though. Dane compared it to measuring gym membership: you know how many people walked through the door, but “it doesn’t tell you whether you’re getting fitter or healthier.”

What matters is what people do with the time that AI frees up. If it goes to strategic thinking, relationship building or even self-improvement, then it's a win. Otherwise, Dane warned, “it just leads to a faster treadmill, with people running more than before.”

So what mindset should we adopt? For Cher Whee, it comes down to a growth mindset – and raising the bar with every interaction. “If you are asking a prompt today, tomorrow you should ask a more ambitious goal of the model. You train the model, but you also stretch your thinking.”

Her advice for getting started is refreshingly practical: start small for quick wins, find a learning buddy or mentor, and celebrate even the small successes. And experience still counts, Dane reminds us: “You’ve done the reps. You’ve been in the trenches 10, 20, 30 years. That amounts to something.”

Food for thought indeed.

Watch the podcast here.