McKinsey sees AI boosting jobs like software development

Sponsored by

Header: 59% of professionals spend 11+ hours a week chasing information across systems.†

Copy: This guide breaks down how to connect context across tools and keep work moving.

†Quickbase, 2025

There’s a lot of AI pessimism out there, but at least one analyst firm sees a bright side: the jobs outlook will improve.

“We think there’ll be more jobs in 2035 than there are today,” Eric Kutcher, senior partner and chair for North America at McKinsey, told a crowd at the consulting firm’s media day in New York on October 7.

Kutcher was referring to the consultancy’s recent report, released on September 29, which estimated:

• Automation could “reduce labor demand by the equivalent of roughly 36 million jobs.”
• AI value could “generate demand for more than 40 million jobs over the next decade.”

Does that mean IT jobs? Many of those future occupations, the report noted, will arise from AI workflows, in areas like model development, infrastructure, agent engineering, workflow design, governance, and human oversight.

“We estimate that AI could create roughly 500,000 to two million new jobs in occupations that do not yet exist or are too small to measure today,” the report authors wrote.

AI company Anthropic, in its own March study, revealed occupations with the most exposure to AI, including computer programmers, data entry keyers, software quality assurance analysts and testers, and information security analysts. The US Bureau of Labor Statistics (BLS) projects occupations with higher exposure will grow less through 2034, Anthropic shared in its report’s key findings.

Exposed but not erased. Kutcher’s view of AI’s influence on the job market appears more optimistic than the Anthropic report’s take. He sees software developers experiencing plenty of exposure to AI, leading to higher demand than a year ago. “What’s going on is the job of that software developer is profoundly different,” Kutcher told attendees. “Being a great software developer doesn’t mean you have to type.”

Recent reports released from Morgan Stanley and the World Economic Forum suggested that, with the advent of AI, developers’ tasks have moved from routine coding to architecture and integration strategy.

James Norman, CEO at market research platform Pilotly and managing partner at Black Operator Ventures, envisions engineers who want to enable AI to solve business-specific problems need to have foundational knowledge in infrastructure like cloud storage buckets, frontier models, harnesses, and agents.

“Today, most organizations don’t have fully adequate engineering teams. They have IT teams. They don’t have engineering teams that build systems for them. They mainly procure software,” Norman told us. “In the future, because organizations are going to operate on central intelligence systems and leverage various AI tools and connectors that plug into them, you need to have a team of people that actually builds infrastructure that allows your entire organization to operate on AI. And that really can’t be outsourced in the end.”

Those architects aren’t necessarily the people who know how to write Python. “It’s the person who’s like, ‘I’m creative, I’m curious, I’m thoughtful about how I solve problems, and I’m also familiar with the systems that make all this stuff work in the first place,’” he said.

Stack Overflow’s developer survey of over 30,000 respondents, released on Oct. 6, found that 62% of software pros view AI in a “favorable” light. Almost 8 in 10 developers are focusing on AI skills as part of their learning regimen.

Daniel Swan, senior partner and global co-leader of tech and AI at McKinsey, said many clients are hiring more software developers than ever, but tasks now include priorities like training agents or helping colleagues without a technical background use software and technology. (He also noted a “massive influx in demand” for cybersecurity professionals.)

“I think one of the things about AI that changes is everybody in the organization has to have a little bit of a software developer mindset, and very few of them actually do,” Swan told us at the event.