Appendix B: Source Index and Further Reading

This source index collects the research spine behind AI-Native: labor economics for task movement, software studies for AI-assisted development, human factors for oversight limits, eval resources for acceptance, and governance sources for responsibility. The sources do not all make the same claim. That is the point.

AI productivity evidence is jagged. Support studies, coding experiments, governance frameworks, and human-AI interaction guidance each illuminate a different part of the operating model. Read them as constraints on the argument: AI-native work needs task decomposition, evaluation, ownership, boundary setting, and a cadence for learning from failure.

Key Takeaways

• The index groups 34 sources into five operating themes.
• Labor and productivity sources explain why work moves at task level before job titles change.
• Software and eval sources explain why output volume is not the same as accepted change.
• Human factors and governance sources explain why "a human reviews it" is not enough.
• The adoption context sources should be refreshed as model capability, regulation, and enterprise practice change.

Labor, tasks, and productivity

• David Autor, MIT Economics profile: https://economics.mit.edu/people/faculty/david-h-autor
• Autor, Levy, and Murnane, "The Skill Content of Recent Technological Change": https://economics.mit.edu/sites/default/files/publications/The%20Skill%20Content%20of%20Recent%20Technological%20Change.pdf
• Brynjolfsson, Mitchell, and Rock, "What Can Machines Learn, and What Does It Mean for Occupations and the Economy?": https://www.aeaweb.org/articles?id=10.1257/pandp.20181019
• Brynjolfsson, Li, and Raymond, "Generative AI at Work": https://www.nber.org/papers/w31161
• Dell'Acqua et al., "Navigating the Jagged Technological Frontier": https://www.hbs.edu/faculty/Pages/item.aspx?num=64700
• BCG summary of jagged frontier field experiment: https://www.bcg.com/publications/2023/how-people-create-and-destroy-value-with-gen-ai
• Erik Brynjolfsson, "The Turing Trap": https://digitaleconomy.stanford.edu/news/the-turing-trap-the-promise-peril-of-human-like-artificial-intelligence/
• The Turing Trap on arXiv: https://arxiv.org/abs/2201.04200

Software work and AI-assisted development

• Peng et al., "The Impact of AI on Developer Productivity": https://arxiv.org/abs/2302.06590
• GitHub research summary on Copilot productivity and happiness: https://github.blog/news-insights/research/research-quantifying-github-copilots-impact-on-developer-productivity-and-happiness/
• METR, "Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity": https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/
• METR paper on arXiv: https://arxiv.org/abs/2507.09089
• SWE-bench official site: https://www.swebench.com/
• DORA AI resources: https://dora.dev/ai/
• 2025 DORA State of AI-Assisted Software Development: https://cloud.google.com/resources/content/2025-dora-ai-assisted-software-development-report

Human factors, decision-making, and human-AI interaction

• Parasuraman and Riley, "Humans and Automation: Use, Misuse, Disuse, Abuse": https://doi.org/10.1177/001872089703900303
• Microsoft Research, Guidelines for Human-AI Interaction: https://www.microsoft.com/en-us/research/project/guidelines-for-human-ai-interaction/
• Evaluating Human-AI Collaboration review: https://arxiv.org/html/2407.19098v1
• Microsoft Research, "Goals as First-Class Abstractions in Human-AI Collaboration": https://www.microsoft.com/en-us/research/publication/goals-as-first-class-abstractions-in-human-ai-collaboration/
• Microsoft New Future of Work Report 2025: https://www.microsoft.com/en-us/research/wp-content/uploads/2025/12/New-Future-Of-Work-Report-2025.pdf

AI systems, agents, and evals

• OpenAI evaluation best practices: https://developers.openai.com/api/docs/guides/evaluation-best-practices
• OpenAI evals guide: https://developers.openai.com/api/docs/guides/evals
• Anthropic, "Building Effective AI Agents": https://www.anthropic.com/research/building-effective-agents
• OpenAI Evals repository: https://github.com/openai/evals

AI governance, risk, and security

• NIST AI Risk Management Framework: https://www.nist.gov/itl/ai-risk-management-framework
• NIST AI RMF 1.0 PDF: https://nvlpubs.nist.gov/nistpubs/ai/nist.ai.100-1.pdf
• OECD AI Principles: https://oecd.ai/en/ai-principles
• EU AI Act Article 14: https://artificialintelligenceact.eu/article/14/
• EU AI Act service desk, Article 14 summary: https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-14
• ISO/IEC 42001:2023 AI management systems: https://www.iso.org/standard/42001
• OWASP Top 10 for Large Language Model Applications: https://owasp.org/www-project-top-10-for-large-language-model-applications/

AI adoption context

• Stanford HAI AI Index: https://hai.stanford.edu/ai-index
• Stanford AI Index Report 2025: https://hai.stanford.edu/ai-index/2025-ai-index-report
• Stanford AI Index Report 2026: https://hai.stanford.edu/ai-index/2026-ai-index-report

Use this page as a research map, not a proof pile. For workflow design, start with the chapters on the work moving inside tasks, acceptance bottlenecks, and the AI-native operating system.