AI, Jobs, and the Skills Bet: What HR Needs to Hear From Palantir's Alex Karp
At the World Economic Forum in Davos, Palantir CEO Alex Karp didn't mince words: he believes AI will "destroy humanities jobs" and that traditional humanities learning will be "hard to market." Whether you agree or not, HR leaders can't ignore the signal. Your hiring model, skills strategy, and internal mobility programs are now business-critical.
What Karp actually said - and why it matters
Karp has long been skeptical of higher education's payoff, even noting his own struggle to get hired after a PhD in philosophy: "I'm not sure who's going to give me my first job." He argues today's workforce needs "different ways of testing aptitude," because older filters don't surface what makes someone irreplaceable.
He's consistent inside Palantir. The company's Meritocracy Fellowship pays high school graduates to study philosophy and history while working on real projects - a direct bet on aptitude and agency over credentials. In Palantir's words: people with the highest aptitude deserve "challenges… agency… and responsibility."
The broader market picture
Karp's take is blunt and likely oversimplifies the full impact, but the direction is clear: AI is changing work. A 2025 Forbes survey found 44% of CHROs have already transitioned employees into AI-related roles. More importantly, 68% are prioritizing human-AI collaboration over replacement.
The same survey shows 94% expect fewer than 5% of roles to be eliminated across 2026-2027. A 2026 Deloitte survey adds a useful lens: high-performing organizations double down on human skills like divergent thinking and curiosity to make teams adaptable and effective with new tech.
Reality check from the frontline
Doug McMillon, former Walmart CEO, put it plainly: "AI is going to change literally every job." His advice is to keep conversations transparent, frequent, and grounded in what you're learning and why you're doing it. That attitude wins trust and buys time for reskilling.
Action plan for HR leaders
- Redesign jobs into tasks: Map roles into task portfolios. Flag tasks for automate, augment, or keep human-led. Rebuild job descriptions around outcomes, tools, and decision rights.
- Shift from pedigree to aptitude: Introduce work-sample tests, scenario-based prompts, and time-boxed build challenges. Reduce overreliance on degrees.
- Create an internal skills graph: Stand up a clean skills taxonomy. Tag employees, tasks, and projects. Use it to route people to high-value work and target upskilling.
- Launch role-to-role transitions: Fund fast tracks into AI-adjacent roles (analyst, prompt engineer, automation ops, QA for AI outputs). Pair with mentors and 90-day milestones.
- Train for human advantage: Prioritize skills AI amplifies: problem framing, critical thinking, domain judgment, creativity, communication, and ethical reasoning.
- Update performance and pay: Reward adoption, outcomes with AI, and cross-functional impact. Make tool proficiency and data quality part of the scorecard.
- Stand up governance: Clear policies for data privacy, accuracy reviews, and human-in-the-loop checkpoints. Assign owners. Audit quarterly.
- Communicate like a product team: Ship AI updates in small releases. Share what changed, who benefits, and the next experiment. Make it two-way.
- Track hard metrics: Time-to-complete, quality, error rates, customer NPS, and employee adoption. Tie training to measurable performance lifts.
Hiring: what to screen for now
- Aptitude over pedigree: Problem decomposition, fast learning loops, and curiosity.
- Tool fluency: Ability to evaluate, prompt, and QA AI tools in real workflows.
- Decision judgment: Knowing when to trust or override AI outputs.
- Collaboration: Can they co-create with humans and systems without ego?
Signals from Palantir's approach
Expect more employers to test for thinking quality, not just credentials. Short, real-world challenges will matter more than resumes. Programs blending humanities with technical execution will produce versatile operators - the people who turn AI into outcomes.
Resources to accelerate upskilling
- AI courses by job role - quick paths for HR, operations, finance, and more.
- Latest AI courses - stay current with tools and practical workflows.
Bottom line
AI will change how value gets created. Whether it "destroys" certain jobs or not, HR's mandate is clear: refactor work into tasks, hire for aptitude, build human skills that compound with AI, and communicate every step with honesty.
Move first. The organizations that learn fastest will keep their people - and their edge.
Your membership also unlocks:
