The Journey from Burnout to Breakthrough
Before his latest AI venture, Steinberger was renowned for creating PSPDFKit, the powerhouse PDF engine used by tech giants like Adobe, Dropbox, DocuSign, and SAP. After ten years at the helm, he sold the majority of his shares in 2021 and took a three-year hiatus to recover from intense burnout.
When he returned to coding in 2024, he found a world transformed by Artificial Intelligence. For Steinberger, AI didn't just automate tasks; it made coding "fun again."
Managing an Army of AI Agents
Steinberger's current workflow is a glimpse into the future of high-level engineering. He operates with 5 to 10 AI agents working simultaneously. Each agent tackles specific tasks and submits a Pull Request (PR).
Remarkably, Steinberger no longer reads the raw code within these PRs. Instead, his focus has shifted entirely to high-level architecture. His system relies on:
AI-Generated Tests: Agents are tasked with writing their own test suites before the actual code.
Automated Linting: Using automated tools to ensure code quality and style.
Bypassing Traditional CI: He has largely moved away from heavy reliance on Continuous Integration (CI) pipelines, arguing that if an AI agent has already validated the code through its own internal testing, it is ready for deployment.
The "30% Rule" for Future Engineering
Steinberger’s conclusion is a wake-up call for the industry: he believes that with AI, he can achieve the same results with only 30% of the headcount previously required. However, there is a catch—those remaining developers must be high-level engineers capable of managing complexity and architecture rather than just writing syntax.
- In the past, software engineers spent most of their time focusing on "how to write code." But in Steinberger's era, the emphasis has shifted to "what to build" and "how to assemble the components" (architecture), skills that AI still doesn't excel at as well as humans.
- Steinberger's disregard for traditional CI doesn't mean he's neglecting it; rather, he utilizes AI's "pre-commit validation," which is faster than waiting minutes for a CI server to run. This represents a shift from "post-commit validation" to "in-process validation."
- The statement, "We only need 30% of engineers, but they must be highly skilled," reflects a major problem facing the industry in 2026: junior developers are rapidly being replaced by AI. This forces new graduates to prioritize learning system design over simply memorizing programming syntax.
- Steinberger proved that AI can solve burnout by eliminating "routine" tasks (grinds) and retaining only "creativity," the aspect software developers love most.
Cloudflare Open-Sources "Moltworker": Bringing Moltbot to the Edge Without the Need for a Mac Mini
Source - The Pragmatic Engineer
No comments:
Post a Comment