> Markdown version of [/videos/1148-you-are-not-an-ai-developer?t=1088](https://www.wearedevelopers.com/videos/1148-you-are-not-an-ai-developer?t=1088). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# You are not an AI developer

Building custom AI models from scratch is usually a futile exercise. True innovation happens when grounded software engineers treat off-the-shelf AI frameworks as composable building blocks.

- **Speakers:** [Zan Markan](https://www.wearedevelopers.com/@zan-markan)
- **Event:** World Congress 2024
- **Published:** August 20, 2024
- **Duration:** 23:55
- **URL:** https://www.wearedevelopers.com/videos/1148-you-are-not-an-ai-developer

## Summary

Software development is historically built on the concept of determinism—reproducible environments, continuous delivery, and infrastructure as code—yet developers constantly manage unpredictable production edge cases through observability. Contrastingly, artificial intelligence introduces an inherently non-deterministic paradigm where programs infer rules directly from vast datasets rather than following explicit human instructions. Drawing parallels between classic pop culture technology and modern large language models, the narrative illustrates how concepts like prompt engineering and hallucination management resemble familiar sci-fi tropes while demanding entirely new application testing strategies.

A crucial insight for technical teams is that the impulse to build custom AI models from scratch is almost always a "yak shaving exercise." Competing with massive cloud providers who burn immense compute resources to train foundation models is futile for the vast majority of organizations. Instead, true business value emerges when teams treat off-the-shelf AI frameworks, such as pre-trained LLM APIs or localized deployments, as composable building blocks. By integrating these existing capabilities into standard software systems, developers bypass overwhelming infrastructure barriers and focus squarely on tangible product innovation.

Much like the smartphone transition successfully redefined mobile interaction, AI currently acts as a profound user experience paradigm shift, enabling conversational and intent-driven interfaces for complex workflows like CI/CD pipeline analytics. Navigating this industry shift requires grounded software developers, not necessarily specialized AI researchers. By leaning heavily into foundational engineering disciplines—prioritizing code maintainability, rigorous automated testing, and resilient operational design—practitioners can successfully bridge the structural gap between predictable system architectures and non-deterministic AI features.

**Keywords:** software determinism challenges, off-the-shelf LLM integration, prompt engineering fundamentals, software observability tools, conversational user interfaces, CI/CD pipeline analytics, local-first mesh networks, foundation model training costs, AI user experience paradigm, non-deterministic software execution, LLM API integration, infrastructure as code environments, yak shaving software development, local LLM deployment, generative AI application architecture

## Chapters

1. **Artificial intelligence representations in popular culture and science fiction** (00:30) — How classic science fiction movies explore human-like attributes and unpredictable behaviors in artificial intelligence.
1. **Building local first data synchronization across transport meshes** (03:37) — Creating offline peer-to-peer applications by syncing data locally without cloud dependencies.
1. **Managing determinism and non-determinism in software development workflows** (04:30) — Why traditional code is inherently reproducible despite the unpredictable nature of real-world user behavior and infrastructure.
1. **Rule inference and logic generation in machine learning models** (09:13) — How machine learning infers rules from training data rather than relying on programmer-defined specific business logic.
1. **Integrating language models into broader established software ecosystems** (12:08) — Why newly trained large language models must be integrated into a broader software architecture to deliver concrete value.
1. **Compute barriers and costs of training custom language models** (14:38) — Immense compute requirements and resource costs make training language models unfeasible for most individual software developers.
1. **Enhancing user experiences with existing artificial intelligence interfaces** (18:08) — Leveraging existing application programming interfaces to add non-deterministic interactions and improve user engagement within current products.
1. **Applying core engineering principles to non-deterministic tooling contexts** (22:13) — Focusing on application maintainability, testing, and system deployment when introducing non-deterministic functional elements into stable engineering practices.

## Related Moments

- [The future role of developers orchestrating artificial intelligence](https://www.wearedevelopers.com/videos/1942-technical-debt-when-vibe-coding) (from "Technical Debt when Vibe coding")
- [Rethinking software engineering processes beyond human constraints](https://www.wearedevelopers.com/videos/1830-wearedevelopers-live-speculaitions) (from "WeAreDevelopers LIVE - SpeculAItions")
- [Introduction to artificial intelligence driven development](https://www.wearedevelopers.com/videos/347-mlops-and-ai-driven-development) (from "MLOps and AI Driven Development")
- [Summarizing developer experience and artificial intelligence companions](https://www.wearedevelopers.com/videos/884-forget-developer-platforms-think-developer-productivity) (from "Forget Developer Platforms, Think Developer Productivity!")
- [Transitioning software engineering teams to AI-native development workflows](https://www.wearedevelopers.com/videos/100087-ai-ready-what-enterprise-transformation-actually-takes) (from "AI-Ready? What Enterprise Transformation Actually Takes")
- [Redefining the software architect role for AI pipelines](https://www.wearedevelopers.com/videos/100190-architecture-3-0-from-90-to-99-999-reliability-in-building-ai-systems) (from "Architecture 3.0: From 90% to 99.999% Reliability in Building AI Systems")

## Related Articles

- [Exploring AI: Opportunities and Risks for Developers](https://www.wearedevelopers.com/magazine/522-exploring-ai-opportunities-and-risks-for-developers)
- [MLOps And AI Driven Development](https://www.wearedevelopers.com/magazine/82-mlops-and-ai-driven-development)
- [Navigating the AI Shift](https://www.wearedevelopers.com/magazine/629-navigating-the-ai-shift)
- [What is Software Engineering in the Age of AI?](https://www.wearedevelopers.com/magazine/640-what-is-software-engineering-in-the-age-of-ai)

## Related Jobs

- [Principal Software Engineer, Enterprise AI Platform](https://www.wearedevelopers.com/jobs/ext/1467292-principal-software-engineer-enterprise-ai-platform) at **GitHub**
- [Staff Software Engineer, Copilot Experiences](https://www.wearedevelopers.com/jobs/ext/164361-staff-software-engineer-copilot-experiences) at **GitHub**
- [AI Software Engineer (Germany)](https://www.wearedevelopers.com/jobs/48317-ai-software-engineer-germany) at **Sunhat**
- [Machine Learning Engineer](https://www.wearedevelopers.com/jobs/ext/588393-machine-learning-engineer) at **Twilio**
- [Machine Learning Engineer](https://www.wearedevelopers.com/jobs/ext/1355348-machine-learning-engineer) at **TWILIO**
- [Head of AI Applications](https://www.wearedevelopers.com/jobs/ext/1456210-head-of-ai-applications) at **ZEISS Group**