Will AI create or take away more job opportunities?

This is one of the most debated questions about AI's economic impact, and the honest answer is that experts disagree significantly, with recent evidence pointing to a complex, uneven picture rather than a simple "more" or "less." Let me search for the latest data on this.## Short answer: Most major forecasts say more jobs overall — but the transition is messy, uneven, and painful for many workers

The data paints a genuinely mixed picture depending on timeframe and what's being measured:

The optimistic, net-positive case

The World Economic Forum's Future of Jobs Report 2025, based on a survey of over 10,000 executives globally, is the most-cited projection: the labour market is expected to experience substantial redistribution rather than simple contraction, with employers projecting 170 million jobs created and 92 million displaced by 2030, resulting in net growth of 78 million jobs.

Similarly, McKinsey's newer analysis of the U.S. finds that AI and automation will cut demand for about 36 million U.S. jobs by 2035 while growth elsewhere creates about 41 million — a net gain, but with a catch: the jobs will exist, but it's getting workers into them that's the problem, with "the next decade's challenge is mobility, not scarcity." Specifically, about 11 million workers, roughly 7% of the workforce, would need to leave their occupations entirely to access the new roles.

The "right now" picture looks more negative

Short-term, real-time data is less rosy. In the short term, AI is not yet creating more jobs than it's destroying — through August 2026, AI was the leading stated reason for U.S. layoff announcements, with 54% of all layoff events citing AI or automation, and research suggests AI will displace more jobs than it directly creates through 2028.

Goldman Sachs' estimate is similarly cautious: roughly 16,000 net U.S. jobs are being eliminated per month — about 25,000 positions automated, partially offset by 9,000 created through augmentation, annualizing to roughly 192,000 net positions, about 0.1% of the workforce.

Why the numbers seem contradictory — they're measuring different things

As one analysis put it: these numbers aren't contradictory, they measure different slices of the same reality — Goldman measures net displacement across the whole economy, while other sources measure new AI-specific roles. The clearer trend: destruction is concentrated at entry level while creation favors experienced workers with AI skills.

The real risk isn't total jobs — it's who gets hurt

This is the consensus concern across nearly every source. The WEF itself flags that entry-level and white-collar roles are most exposed: while 170 million new jobs are projected this decade, AI-powered tools threaten to automate as many roles as they create, particularly for entry-level roles — Bloomberg finds AI could replace more than 50% of tasks performed by market research analysts (53%) and sales reps (67%), versus just 9-21% for managers. Overall, 40% of employers expect to reduce their workforce where AI can automate tasks.

A granular U.S. breakdown of occupational exposure found: only 10% to 15% of US jobs face elimination, while 50-55% will change substantially but survive — between three and five jobs get reshaped for every one that disappears.

My take

If you're asking "will there be enough jobs in total?" — most serious forecasts (WEF, McKinsey) say yes, with net gains likely over a 5-10 year horizon.

If you're asking "will my job be fine?" — that depends heavily on:

So the honest synthesis: AI will likely create more jobs than it destroys in aggregate, but it will destroy specific jobs faster than it creates the specific new ones, and the people losing jobs are frequently not the same people getting the new ones. That mismatch — not a shortage of jobs — is the real challenge experts are now focused on.