According to Fortune, LinkedIn’s data presents a counter-narrative to pure AI job loss panic, stating that “in the near term, AI is creating more jobs than it is replacing.” Their report highlights a new category of “new collar” jobs, with about 1.3 million such roles added globally between 2023 and 2025. These include positions like forward-deployed engineers, data annotators, and forensic analysts. Sue Duke, LinkedIn’s Head of Global Public Policy, notes that hiring is otherwise sluggish, about 20% below pre-pandemic levels in advanced economies. She also warns that a staggering 70% of the skills needed for the average job will have changed by 2030, forcing a major shift in how we evaluate candidates away from “legacy signals” like education pedigree.
The bright spot that hides a storm
So, AI is a job creator right now. That’s the good news, and in a slow hiring market, any growth is welcome. But here’s the thing: focusing on these shiny new job titles—Head of AI, AI Engineer—feels a bit like rearranging deck chairs. The real story isn’t the new roles; it’s the fundamental rewiring of almost every existing role. 70% skills change by 2030 is an insane number. That’s not evolution; that’s a forced revolution.
And that shift is already causing chaos in hiring. The number of applicants per job has doubled since 2022. Companies use AI to sift resumes; candidates use AI to game those systems. It’s a dispiriting arms race that often misses the point. When Duke says we need to move from “what school did you go to?” to “do you have the skills and potential?”, it sounds great. But practically, that’s incredibly hard. How do you measure “potential” at scale? Legacy markers were a lazy heuristic, but we’re struggling to find a better one.
Human in the lead, not just the loop
The article ends with a crucial distinction: we need humans “in the lead,” not just “in the loop.” This is the real insight. You can’t automate trust, judgment, or strategic vision. The soft skills gap might become the biggest chasm of all. Think about it. The tech engineer who can’t explain their work to a non-technical executive, or the executive who gets lost when the water-cooler chat turns to token sequences? They’re both at risk.
Basically, we’re looking at a workforce bifurcation. On one side, you have the “new collar” builders and handlers of the AI stack. On the other, you have everyone else whose job is being subtly—or not so subtly—augmented and altered by the tools those new collar workers build. The companies that win will be the ones who figure out how to bridge that gap internally with relentless, continuous reskilling. The ones that lose will just keep chasing the latest job title to drop in the chat.
