This is one of the most debated questions in economics and technology right now, and the honest answer is: it depends on the timeframe, the industry, and who you ask—there's no consensus even among experts. Rather than search for a definitive "answer" (which doesn't exist), let me share the key perspectives on both sides so you can weigh them yourself.
The case for AI taking away more jobs
•Automation of routine cognitive work: Unlike previous waves of automation that mainly affected manual/physical labor, AI (especially generative AI) can now do tasks previously thought safe—writing, analysis, coding, customer service, basic design, and even some legal/medical triage work.
•Speed of disruption: Previous technological shifts (industrial revolution, computers) played out over decades, giving labor markets time to adjust. Some economists worry AI's capabilities are advancing faster than new job categories can emerge or than people can retrain.
•Winner-take-most dynamics: A small number of highly skilled workers augmented by AI may be able to do the work of many, reducing headcount needs in some fields (translation, basic content creation, entry-level coding, paralegal work).
The case for AI creating more jobs
•Historical pattern: Every major technology (the printing press, electricity, computers, the internet) initially destroyed certain jobs but ultimately created more new ones than it eliminated—often in roles that didn't exist before the technology did (e.g., "social media manager" or "app developer" would have been meaningless job titles 30 years ago).
•New categories emerging already: Prompt engineering, AI ethics/safety roles, AI auditing, data labeling, human-AI collaboration specialists, and oversight roles are growing.
•Productivity-driven growth: When AI makes companies more productive and profitable, that often leads to expansion, new product lines, and new hiring—not just cost-cutting.
•Augmentation vs. replacement: Many economists argue AI will augment most jobs rather than fully automate them, changing how work is done rather than eliminating the need for humans.
Why it's genuinely uncertain
•Timeframe matters a lot: Short-term (next 5 years) displacement in specific sectors (customer service, copywriting, basic coding, translation) seems likely even to optimists. Long-term (20+ years) net effects are much harder to predict.
•It depends on policy choices: How governments handle retraining, education, safety nets, and regulation will heavily influence whether displaced workers land in new roles or fall into prolonged unemployment.
•Quality of jobs matters, not just quantity: Even if total job numbers stay flat or grow, the types of jobs, wages, and required skills could shift dramatically, creating real hardship even in a "net positive" scenario.
•Geographic and demographic unevenness: Effects won't be uniform—certain industries, regions, and skill levels will be hit much harder than others.
Bottom line: There's no confident consensus. Serious economists, labor researchers, and tech leaders land on different sides depending on their assumptions about the pace of AI capability growth, policy responses, and how quickly new industries emerge. If you want, I can look into specific recent forecasts or studies (e.g., from McKinsey, WEF, or specific economists) that try to quantify this—let me know if that would help.