Ryan Roslansky has spent the last several years watching artificial intelligence reshape how people work, hire, and get hired. Now the LinkedIn CEO is making a prediction that might surprise the engineers and data scientists racing to future-proof their careers: the most durable professional advantage in an AI-saturated economy won’t be technical. It’ll be human.
In a recent interview with Business Insider, Roslansky argued that soft skills — communication, leadership, adaptability, creative problem-solving — are becoming more valuable, not less, as AI automates an expanding share of routine cognitive tasks. His reasoning is straightforward. If AI can write code, draft legal briefs, and generate financial models, then the people who rise will be those who can do what machines still can’t: persuade a boardroom, manage a fractious team through a crisis, or read the room well enough to close a deal that looked dead.
“The skills that are going to matter most are the ones that are hardest to automate,” Roslansky told Business Insider.
That’s not a particularly novel observation in isolation. Management consultants and futurists have been saying some version of it for years. But coming from the person who runs the world’s largest professional network — with more than a billion members and granular data on hiring trends across virtually every industry — the statement carries different weight. LinkedIn doesn’t just observe the labor market. It is the labor market’s central nervous system, the platform where recruiters post jobs, candidates signal availability, and employers increasingly rely on skills-based matching algorithms to find talent.
And the data backs Roslansky up. LinkedIn’s own workforce reports show that the skills required for the average job have changed roughly 25% since 2015. By 2030, the company projects that figure will reach 65%. Much of that shift is being driven by AI, which is simultaneously creating new technical roles and hollowing out tasks that once defined existing ones. When the task layer gets automated, what remains is the judgment layer — and judgment is built on soft skills.
Roslansky isn’t alone in this assessment. A recent wave of commentary from CEOs, economists, and AI researchers has converged on a similar conclusion, though with varying degrees of urgency. Anthropic CEO Dario Amodei has spoken publicly about how AI will likely handle most routine analytical work within a few years, pushing the premium toward interpersonal and strategic capabilities. OpenAI’s Sam Altman has made parallel observations about the declining half-life of purely technical knowledge.
The timing matters. Companies are in the middle of the largest AI deployment cycle in corporate history. According to a May 2025 report from McKinsey, more than 70% of large enterprises have adopted generative AI in at least one business function, up from 33% just 18 months earlier. That adoption is creating a paradox in the labor market: demand for AI literacy is surging even as the tools themselves are becoming easier to use. The implication is that knowing how to prompt an AI model will soon be table stakes — about as differentiating as knowing how to use Excel. What will separate candidates is everything else.
LinkedIn has been positioning itself accordingly. The platform has invested heavily in AI-powered features — an AI assistant for job seekers, AI-generated drafts for recruiters, algorithmic matching that weights skills over traditional credentials like degrees and job titles. But Roslansky has also pushed the company to surface and validate soft skills more explicitly in user profiles and job postings. It’s a bet that the market will increasingly price these capabilities.
Some skeptics argue that “soft skills” is too vague a category to be operationally useful. Fair point. The term encompasses everything from emotional intelligence to public speaking to conflict resolution. Not all of these are equally valuable in every context, and measuring them reliably remains notoriously difficult. A LinkedIn endorsement for “leadership” tells a recruiter almost nothing.
Roslansky seems aware of this limitation. In his Business Insider interview, he emphasized the need for better ways to assess and credential these abilities — not just list them. LinkedIn has been experimenting with skills assessments and verified credentials, though the platform hasn’t yet cracked the code on validating something as inherently subjective as “adaptability” or “creative thinking.”
Still, the direction of travel is clear. Employers are already adjusting. A growing number of Fortune 500 companies have adopted skills-based hiring practices that de-emphasize four-year degree requirements in favor of demonstrated competencies. Walmart, Google, IBM, and Accenture have all made high-profile moves in this direction over the past two years. And when these companies talk about the skills they’re prioritizing, soft skills consistently rank near the top — often above specific technical certifications.
The education sector is scrambling to catch up. Business schools, once focused almost exclusively on quantitative rigor, are expanding curricula around collaboration, ethical reasoning, and cross-cultural communication. Stanford’s d.school, Harvard Business School’s field method, and similar programs at Wharton and INSEAD all reflect a recognition that analytical horsepower alone won’t cut it when AI can match or exceed human performance on many analytical tasks.
There’s a labor economics angle here too. David Autor, the MIT economist whose research on automation and employment is among the most cited in the field, has argued that AI will likely increase the value of “new work” — tasks and roles that don’t yet exist but will emerge as humans and machines find new ways to collaborate. History suggests these new roles will disproportionately reward social intelligence, creativity, and the ability to synthesize information across domains. In other words, soft skills by another name.
Not everyone is sanguine about this transition. Critics point out that the workers most vulnerable to AI displacement — those in administrative, clerical, and entry-level analytical roles — are often the same workers who have the fewest resources to invest in developing higher-order interpersonal skills. A customer service representative whose job is automated by a chatbot doesn’t automatically become a skilled negotiator or team leader. The gap between “soft skills matter more” and “here’s how displaced workers actually acquire them” remains vast.
Roslansky’s LinkedIn is trying to address this in part through its LinkedIn Learning platform, which offers courses on everything from emotional intelligence to executive presence. The company reported in early 2025 that enrollment in soft-skills courses had increased more than 100% year over year, with particularly strong growth in modules related to AI-era leadership and communication. Whether online courses can genuinely build these capabilities — as opposed to merely introducing concepts — is an open question.
What isn’t debatable is the speed at which AI is changing the composition of work. A May 2025 analysis by the World Economic Forum estimated that 44% of workers’ core skills will be disrupted over the next five years, with analytical thinking, creative thinking, and resilience topping the list of skills expected to grow in importance. Technical skills like programming and data analysis ranked lower — not because they’re unimportant, but because AI tools are increasingly handling them.
So where does this leave the average professional? Roslansky’s advice, distilled, is pragmatic: invest in the skills that compound over a career rather than the ones that depreciate with each new model release. Learn to use AI tools, yes. But don’t mistake tool proficiency for strategic value. The person who can orchestrate a team, communicate a vision, and adapt when the plan falls apart will outlast the person who can write the best prompt.
It’s a message that cuts against the prevailing anxiety of the moment, which tends to focus on what AI will take away. Roslansky is making the case for what it can’t.
Whether the labor market will reward that bet as cleanly as LinkedIn’s CEO suggests remains to be seen. Markets are messy. Transitions are uneven. And the history of technological disruption is littered with confident predictions that aged poorly. But the underlying logic — that as machines get better at thinking, humans need to get better at everything else — is hard to argue with.
And LinkedIn, with its unmatched view of how a billion professionals present themselves to the world, is positioned to shape the answer as much as predict it.
WebProNews is an iEntry Publication