The data is clear on what’s already happening.
Labor’s share of GDP fell to 53.8% in Q3 2025 — the lowest in the modern BLS series back to 1947. Capital is eating labor’s lunch, and AI is accelerating it.
Entry-level pressure is already visible: Stanford’s Digital Economy Lab found declines concentrated among 22–25 year-old workers in AI-exposed jobs such as software development, customer service, and clerical work.
The narrower BLS category of computer programmers is projected to decline 6% through 2034, even while broader software development roles still grow. That distinction matters: implementation-heavy work gets pressured first.
Meanwhile, workers with AI skills command meaningful wage premiums. PwC found an average 56% wage premium for workers with AI skills in 2024, and other labor-market reporting showed premiums up to 43% for jobs listing multiple AI skills.
Tech salary growth slowed to 1.6% in 2025, down from 3.5% in 2023.
This is not speculation. The compression is measurable.
Historical pattern — this has happened before, twice
Web development (late 90s → 2008): Developers charged premium rates to technically naive clients. WordPress, Squarespace, and frameworks killed the implementation premium. Value migrated to architecture, product strategy, integration.
Mobile apps (2008 → mid-2010s): App Store gold rush → React Native / Flutter commoditized a large chunk of standalone app building. Same outcome: implementation became cheaper, judgment became more expensive.
The arc is usually the same: scarcity premium → tool-driven commoditization → value migrates upward to domain expertise, architecture, integration, and judgment.
Historically, that took 5–8 years. AI appears to be compressing the cycle closer to 2–3 years.
What the capital side of this looks like
Piketty’s r > g framework feels increasingly visible in current data.
- The S&P 500 is dramatically above its January 2023 level.
- AI captured close to 50% of all global venture funding in 2025, with roughly $202.3B invested across infrastructure, foundation labs, and applications.
- McKinsey has explicitly warned that when real estate and equity values rise faster than GDP, capital can get pulled toward asset inflation and repurchases rather than the kinds of investment that generate broad long-run growth.
Labor income alone is becoming a weaker wealth-building vehicle. The gap between returns on capital and returns on labor is widening, not narrowing. For someone starting from near-zero capital, that is the fundamental challenge.
The critical prediction for 2026–2030
Middle-class software developer income will likely bifurcate.
- Bottom 60% of current developers: real income stagnation or decline. Implementation work commoditizes.
- Top 20%: premium widens. Architecture, security, compliance, domain-specific systems, AI integration, and judgment-heavy work become more valuable.
- The gap between these groups likely widens through at least 2030.
Not because software disappears.
Because undifferentiated software work gets repriced downward, while the value of high-context, high-trust, high-complexity work rises by contrast.
The uncomfortable truth
Software commoditization means the floor rises — more people can build — but the ceiling also rises. Complex integration, security, compliance, architecture, and domain-specific work become more valuable precisely because raw implementation gets cheaper.
That usually hollows out the middle.
The real risk is getting distracted by opportunities that look easier, when in practice they are just more crowded.
Top comments (0)