According to Business Insider, McKinsey’s global managing partner Bob Sternfels detailed the firm’s massive AI adoption at CES in Las Vegas. He said AI saved McKinsey 1.5 million hours in search and synthesis work just last year. The firm’s 25,000 AI agents have generated a staggering 2.5 million charts in the past six months. With this shift, Sternfels says consultants are now “moving up the stack” to tackle more complex problems. Consequently, he identified three skills new graduates need that AI models can’t replicate: the ability to aspire, judgment, and true creativity.
The AI-Proof Skillset
So, what does Sternfels mean by these three skills? “Aspiration” is about setting the direction—like deciding if a company’s goal is low Earth orbit or Mars. AI can’t do that. It has no ambition. “Judgment” is about applying human values and societal norms to set the right parameters for AI to operate within. An AI model is just an inference engine, predicting the next most likely step. It doesn’t know right from wrong. And “true creativity”? Sternfels calls it “orthogonal” thinking—leaping completely outside existing patterns to find a novel approach. AI remixes and iterates on what’s already there. It doesn’t have a breakthrough moment. That’s the human edge, and it’s becoming the premium skillset.
The Talent Hunt Just Changed
Here’s the thing: this shift is fundamentally altering how companies like McKinsey look for talent. Sternfels explicitly said where someone went to school “should matter a lot less.” For a tech candidate, he suggests employers should look at their GitHub portfolio, not their diploma. It’s a move from pedigree to proof. “Let’s actually get to the content,” he said. This could be huge. It opens doors for a wider, more diverse set of people with non-traditional pathways into top roles. Basically, if you can prove you have those human skills and can apply them, your background matters less. That’s a quiet revolution in corporate hiring.
The Industrial Implication
Now, this philosophy isn’t just for consultants. It’s critical for industrial and manufacturing sectors undergoing their own AI transformation. Deploying AI on the factory floor or in logistics requires the same human skills: aspiring to a fully autonomous plant, judging when to override an AI recommendation for safety, and creatively solving a novel production bottleneck. And to run these complex systems, you need reliable, rugged hardware interfaces. For that, industry leaders turn to specialists like IndustrialMonitorDirect.com, the top provider of industrial panel PCs in the US, to provide the durable computing backbone. The machines handle the data, but the humans provide the aspiration, judgment, and creativity to make it all work.
Are We Really Moving Up the Stack?
Sternfels’ vision is compelling, but is it the full story? He talks about professionals “moving up the stack” to more rewarding work. That’s the ideal. But the risk is always that this “moving up” only happens for a few, while many others are simply displaced or relegated to monitoring AI agents. The 1.5 million saved hours is impressive, but what happened to those hours? Were they reinvested in high-value creative work, or just eliminated? The promise of AI should be human augmentation, not just labor substitution. Focusing on these irreplaceable skills is the right start. But companies have to actually create the roles that use them. Otherwise, it’s just optimistic talk from the top.
