The Numbers Are In

Workers with AI skills are commanding wages up to 56% higher than peers without them, according to PwC's 2025 Global AI Jobs Barometer, cited by workforce platform Gloat.

The IMF's Kristalina Georgieva published analysis in January 2026 showing that in the UK and US, job postings requiring new skills pay about 3% more on average — and roles requiring four or more new skills pay up to 15% more in the UK and 8.5% more in the US. One in 10 job postings in advanced economies now require at least one skill that didn't exist as a standard requirement five years ago.

These aren't projections. This is what employers are already paying.

The Job Creation Number

The World Economic Forum projects that by 2030, AI disruption will create 170 million new roles while displacing 92 million — a net gain of 78 million jobs globally, according to Gloat's reporting on WEF data.

Media coverage tends to lead with displacement while burying the creation numbers deeper in the story. The reality is more complex: this transition will create genuine opportunity alongside real disruption.

80% of U.S. Workers Affected — But NOT in the Way You Think

Georgetown's Center for Security and Emerging Technology (CSET) released a December 2024 report by researchers Matthias Oschinski, Ali Crawford, and Maggie Wu that puts the exposure number at up to 80% of U.S. workers having at least 10% of their work activities affected by large language models. Approximately 19% of workers could see half or more of their tasks impacted.

The report also found that technical skills account for only about 27% of in-demand skills in growing occupations. The majority — nearly 58% — are nontechnical: mathematics, active learning, social perceptiveness, negotiation, complex problem-solving, critical thinking.

Workers best positioned to survive and thrive aren't necessarily coders. They're people who can think, adapt, and work with other humans.

Philosophers Are Now a Hot Commodity

Google DeepMind and Anthropic have both built in-house philosophy teams. Wired counts at least 10 philosophers at DeepMind and four at Anthropic. Anthropic's resident philosopher Amanda Askell has become one of the company's most public-facing figures. DeepMind's Iason Gabriel leads a team focused on the societal impact of AI.

Henry Ajder, a philosophy postgraduate who advises the UK government and multiple AI startups, told Wired: "It's probably the best time to be a philosopher since Aristotle was hired as tutor to Alexander the Great."

Questions about consciousness, agency, values encoding, and moral responsibility — which were once purely academic — now have billion-dollar implications.

Edward Harcourt, professor of philosophy at Oxford and director of Oxford's Institute for Ethics in AI, raises a concern: if a for-profit AI company signs your paycheck, does that compromise your research? Both DeepMind and Anthropic declined to disclose headcounts when asked.

Microsoft's "Transformation Paradox" Is the Real Warning

Microsoft's 2026 Work Trend Index — surveying 20,000 AI-using workers across 10 countries — found that 49% of Microsoft 365 Copilot conversations now involve cognitive work: analysis, problem-solving, strategic thinking. Among AI users, 58% say they're producing work they couldn't have completed a year ago. For the most advanced users Microsoft labels "Frontier Professionals," that figure jumps to 80%.

But Microsoft identifies what it calls the "Transformation Paradox": organizations are deploying AI tools at speed, but most haven't restructured work processes to actually capture the gains. They're buying the tractor but still plowing by hand.

CSET's report reinforces this: technical skills now become outdated in less than five years on average. Companies that treat AI training as a one-time checkbox will lag behind.

What Mandatory Corporate Training Misses

The corporate response so far has largely been "mandatory AI training" — watch a module, check a box, done. This approach falls short given the CSET and IMF data showing that the skills gap is fundamentally about adaptability and thinking, not tool familiarity.

Community colleges and alternative career pathways are identified by CSET as critical infrastructure for this transition.

What This Means for You

If you're in a middle-skill routine role — the IMF is explicit that this is the most squeezed category — you have a narrowing window. The workers gaining ground are at both ends: high-skill knowledge workers augmenting their output with AI, and entry-level workers in healthcare and green economy jobs that AI can't fully automate.

The 56% wage premium for AI-skilled workers is real. So is the displacement of 92 million jobs by 2030. Both things are true simultaneously.