Nvidia CEO Calls AI Job Fears 'Complete Nonsense'
Key insights
- GitHub commits grew from 300 million in 2023 to 500 million in 2025, with output nearly tripling in early 2026.
- Huang calculated that AI-tripled developer productivity converts a $3 trillion salary base into $9 trillion of GDP output.
- Open-source maintainers at Godot and RPCS3 reported low-quality AI-generated pull requests increasing their review workload.
Why this matters
Huang's GitHub commit data gives the AI industry a concrete, named-source rebuttal to displacement narratives, directly influencing how companies justify headcount decisions in a hiring cycle already shaped by AI tool adoption. The countervailing evidence from Godot and RPCS3 maintainers shows that commit volume growth can coexist with rising review burden, meaning productivity metrics and labor demand metrics may be moving in opposite directions. Technical leaders in 2026 now have a high-profile, publicly cited data point on each side: the $9 trillion productivity argument favoring expansion, and the "AI slop" quality problem creating new maintenance costs.
Summary
Jensen Huang declared at GTC Taipei that fears of AI eliminating software engineer jobs are "complete nonsense," pointing to surging GitHub commit data as evidence.
Commits grew from 300 million in 2023 to 500 million in 2025, with output nearly tripling in early 2026. Huang's argument: 30 million developers generate $3 trillion in GDP today; triple their output with AI and you get $9 trillion from the same salary base, making more hiring the rational call.
Essentially: (Nvidia, GitHub) -- Huang frames AI as a force multiplier for developers, not a replacement.
- GitHub commits nearly tripled in early 2026 from a 300 million baseline in 2023.
- Huang put $9 trillion of potential productivity against $3 trillion in developer salaries.
- Open-source teams at Godot and RPCS3 flagged waves of low-quality "AI slop" pull requests as adding maintainer burden.
The Linux kernel now permits AI-generated code but requires human review and developer accountability, a policy that shows even permissive institutions are distinguishing output volume from output quality.
Potential risks and opportunities
Risks
- If the Linux kernel's human-review requirement for AI-generated code proves unscalable, major open-source projects face maintainer burnout as AI-assisted contribution volumes grow without quality gates
- Nvidia and Jensen Huang face credibility risk if software engineer headcounts contract in a 2026 hiring downturn, directly contradicting the more-hiring thesis stated at GTC Taipei
- Open-source project leads at Godot, RPCS3, and similar communities may impose AI contribution restrictions if review burden continues to grow, fragmenting the open-source development ecosystem
Opportunities
- Code review tooling vendors and AI code quality platforms are positioned to build pipeline demand on the back of the review burden surge flagged by Godot and RPCS3 maintainers
- Nvidia and GitHub have a shared interest in co-publishing longitudinal commit quality data alongside volume metrics to strengthen the productivity argument and pre-empt the AI slop counternarrative
- Enterprise software teams can use Huang's $9 trillion productivity framing as internal justification to expand AI tooling budgets while maintaining or growing engineering headcount through 2026
What we don't know yet
- Whether Huang's GitHub commit growth figures account for the surge in low-quality AI-generated commits flagged by Godot and RPCS3 maintainers as adding review burden rather than productive work
- Huang's $3 trillion GDP figure for 30 million developers is cited without a source or growth trajectory, making the productivity multiplier math difficult to independently verify
- No employer-level hiring data is provided for any named company, leaving the claim that software engineer numbers are actually increasing unsubstantiated beyond GitHub commit volume
Originally reported by pcguide.com
Read the original article →Original headline: Jensen Huang Calls AI Job Fears 'Complete Nonsense' at Computex — Says Software Engineer Numbers Are Actually Increasing