The AI Employment Question: Will Software Engineers Still Have Jobs in 2030?

If AI can write code, what happens to coders? On prediction markets, forecasters give essentially 50/50 odds—49%—that there will be more software engineers in five years than there are today. This remarkable uncertainty captures genuine disagreement about AI's impact on the profession that created it.

The Coin-Flip Probability

A 49% probability of employment growth (51% probability of decline or stagnation) represents maximum uncertainty. Forecasters are evenly split, which tells us:

• The question is genuinely hard to predict
• Both outcomes have substantial supporting arguments
• The impact of AI on software engineering is unprecedented
• Anyone claiming to know the answer is overconfident

The Growth Case (49%)

Reasons software engineering employment might increase:

Demand expansion: AI tools make software development faster and cheaper. This could expand the total amount of software produced, requiring more engineers even if each is more productive.

Complement not substitute: Current AI coding assistants augment human work rather than replacing it. Engineers become more productive but remain essential.

Quality and complexity: AI-generated code often requires human oversight, debugging, and integration. Complex systems still need human architects.

New domains: AI enables software in areas previously too expensive to develop. These new applications require engineers.

Historical pattern: Previous automation tools (compilers, IDEs, frameworks) made programming more accessible and increased total employment, not reduced it.

The Decline Case (51%)

Reasons software engineering employment might decrease:

AI capability trajectory: Current coding AI is impressive; future versions may handle more complex tasks independently.

Economic pressure: Companies face strong incentives to reduce labor costs. If AI can do the job, engineers will be replaced.

Junior role elimination: Entry-level coding tasks are most vulnerable to AI. This could shrink the pipeline of new engineers.

Efficiency gains: Even without full replacement, 2x productivity per engineer means half as many engineers needed for the same output.

Structural shift: Software engineering could follow manufacturing—fewer, higher-skilled workers producing more output.

The GDP Impact Question

A related market gives 46% odds of visible AI impact on U.S. GDP, unemployment, or productivity by 2028. This tracks closely with the software engineer question—both reflect uncertainty about whether AI's effects will be transformative soon or take longer to materialize.

If AI doesn't visibly impact GDP by 2028, it probably hasn't eliminated software engineering jobs either. If it does impact GDP, employment effects are likely following.

What This Means

The 49% probability has implications:

For current engineers: Skill development matters more than ever. Specialization in areas AI handles poorly (system design, human-AI collaboration, novel problem-solving) may provide protection.

For students: Entering software engineering is not the safe career bet it was a decade ago—but it's not clearly a mistake either.

For companies: Workforce planning must account for rapidly changing productivity curves. Over-hiring or under-hiring both carry risks.

For policymakers: Labor market disruption in high-skill, high-wage sectors would have different dynamics than historical manufacturing displacement.

The Irony

Software engineers are building the tools that might replace them. The profession that automates other jobs is now automating itself. The 49% probability captures this unique moment: the community closest to AI capabilities is genuinely uncertain about its own future.

Conclusion

At 49%, prediction markets have rendered the most honest possible verdict on AI's impact on software engineering: they don't know. Neither confident predictions of massive job losses nor reassurances of continued growth are supported by forecaster consensus. The next five years will resolve this uncertainty—and software engineers are living through the experiment.

---

Analysis informed by aggregated forecaster data from Manifold Markets as of January 20, 2026.