Back in February 2017, Mark Cuban sat down with Bloomberg's Cory Johnson at the NBA All-Star Technology Summit in New Orleans and said something that still gets quoted today. He argued that in 10 years, liberal arts majors would be in higher demand than programming majors, because software would spit out the data and someone would still need to make sense of it.
That's a longstanding opinion, not a prediction with a due date stamped on it. Cuban himself hedged it, saying liberal arts grads "will starve for a while" before employers caught up to his way of thinking. Nearly a decade later, there's enough real-world data to check his reasoning against what's actually happening in offices and code shops.
If you're mid-career and watching AI creep into your field, the question isn't whether your job title survives. It's whether you're building the kind of judgment that's hard to automate, the same way you'd look for ways to make money from home if your hours got cut.
Cuban's case for thinkers over typists
In that 2017 interview, Cuban described what he called an "automation of automation," where the coding and software jobs doing routine work would eventually get replaced by computers themselves. He singled out accounting-style technical tasks as especially exposed.
His point wasn't that technical skill is worthless. It was that technical tasks performed by rote are easier to hand off to a machine than work that requires personal judgment, critical thinking, or creativity. He specifically called out English, philosophy, and foreign language majors as fields likely to see rising demand because of that distinction.
That's a directional bet about where value shifts, not a claim that any particular degree or skill became obsolete on a fixed timeline.
AI coding assistants are already handling the routine stuff
If you want to see Cuban's automation-of-automation idea in action, look at how programmers actually use AI today. Stack Overflow's 2026 Developer Survey found coding assistants and agents are now the dominant AI use case among developers, at 66%, ahead of general chatbots at 63%, and 73% of users rely on them daily.
But the pattern inside that usage matters. Developers lean on AI heavily for generating familiar code, debugging, and answering straightforward technical questions. The moment the task shifts toward communication or design, usage drops sharply, from more than 70% useful on technical work down to 35% for communication and 49% for design, according to the same survey.
Developers aren't sitting back and letting the tools take over, either. More than half, 52%, learned a new coding skill or programming language in the past year, per The Register's October 2026 reporting on the survey. Even so, 17.1% of respondents said losing job skills was a reason to be cautious about leaning on AI at work.
The split between coding mechanics and software judgment shows up in the job numbers
A Stanford Digital Economy Lab study using ADP payroll data found entry-level software developers aged 22 to 25 saw employment fall nearly 20% from its late-2022 peak by July 2025. Workers 30 and older in those same AI-exposed fields didn't see that drop. In fact, their employment in the highest AI-exposure categories grew between 6% and 12% over roughly the same period.
The researchers described this as automation versus augmentation: younger workers doing codified, routine tasks lost ground, while workers whose jobs leaned on judgment gained it. The U.S. Bureau of Labor Statistics still projects software developer employment overall will grow 17.9% between 2023 and 2033, reasoning that someone still needs to design systems and decide what to build, even as AI handles more of the typing.
Stanford's own researchers are careful here, noting their findings are "consistent with" AI affecting entry-level jobs rather than proof of it. BLS says the same about its own projections.
Employers still want people who can think, not just execute
The World Economic Forum's Future of Jobs Report 2025, published in January 2025, found analytical thinking remains the single most sought-after core skill among employers, with 70% calling it essential. Resilience, flexibility, and agility ranked second at 67%, followed by leadership and social influence at 61% and creative thinking at 57%.
A LinkedIn analysis from March 2025 echoed that, finding creativity, problem-solving, and strategic thinking topped its list of fastest-growing workplace skills for the year. That doesn't mean technical skills are fading, though. The same WEF report found AI and big data skills are growing fastest of all, meaning technical fluency and human judgment are rising together rather than trading places.
More workers are retraining the skills they already have
Coursera's Global Skills Report 2026, published in September 2026, found learners across Coursera and Udemy enrolled in generative-AI content at a combined rate of more than 45 enrollments per minute in 2026, up from 25 per minute the year before. Enrollments in critical-thinking courses, the report's top "human skill" pairing with AI learning, rose 107% year over year.
Micro-credential enrollments climbed 33% overall, with AI-specific micro-credentials up 122%. LinkedIn's 2025 Workplace Learning Report found companies that prioritize career development are 32% more likely to be running AI training programs than those that don't.
Formal training still lags behind the enrollment surge, though. A Pew Research Center survey from October 2024 found that among workers who'd taken any job training in the prior year, only 24% said it involved AI use.
Workers who lean into AI tools report sturdier job security
PwC's 2026 Global Workforce Hopes and Fears Survey, which interviewed nearly 50,000 workers across 48 countries in May and June 2026, found daily AI users report 68% confidence in their job security, compared with 57% among infrequent users. Daily users were also more likely to ask for a promotion, trust top management, and feel confident they could learn new skills.
A June 2026 Gallup analysis found 62% of laid-off workers were people who used AI once a year or less. Among tech workers specifically, infrequent AI users were about three times more likely to have been laid off than regular users. Gallup is careful to note this is a pattern, not proof of cause, since only 1% of laid-off workers actually cited AI or automation as the reason for their job loss.
Bottom line
Cuban's 2017 comments hold up less as a prophecy and more as a pretty good read on direction. The data from coding to the broader workforce keeps pointing the same way: routine, checkable tasks are getting automated fastest, while judgment, analysis, and creative problem-solving are holding their value or gaining it.
The practical move isn't abandoning your technical skills. It's pairing them with the kind of thinking a machine still can't do on its own, and treating ongoing learning as part of the job rather than an extra.
If you're mid-career, that might mean taking a short course on the AI tools already showing up in your field, or simply paying closer attention to the parts of your job that involve judgment calls rather than repetitive steps. It's also worth looking at your broader financial footing while you do it, the same way you'd check in on your plans to set yourself up for retirement before a slow year catches you off guard.
Add Us On Google