Software's centaur age may last decades
Short summary
Human software engineers paired with AI coding agents are currently more effective than either working alone, and this partnership phase could persist for 10-20 years before AI achieves full autonomy — giving engineers time to adapt rather than panic. The author argues we should treat software's trajectory like chess, which had a 20-year centaur age before AI dominance, though the outcome remains genuinely uncertain.
- Current AI coding agents (like Claude and Cursor) are reliable enough to work unsupervised but still need human review for alignment issues — matching technical values, avoiding over/under-engineering — making human-AI teams superior to either alone.
- Chess provides the closest historical parallel: its centaur age lasted about 20 years, while previous centaur ages (like humans with knitting frames) lasted 200+ years, suggesting software's could reasonably span a decade or two.
- The main risk isn't staying in software engineering for another decade — it's refusing to become a centaur by rejecting AI tools, or trying to jump ship early and give up a full career based on worst-case scenarios that may not materialize.
Full summary
We're in a "centaur age" of software engineering where humans paired with AI systems outperform either alone. GitHub Copilot launched this phase in 2022 as AI-powered autocomplete. The tech evolved quickly—chatting with GPT-4 helped with unfamiliar areas, then coding agents like Cursor's agent mode (2024) and Claude Code (early 2025) started handling unsupervised work. By November 2025's Claude Opus 4.5 release, agents became reliable enough to run without constant supervision, though they still need human review. No unassisted engineer can beat a centaur team anymore, but unsupervised agents can't beat them either. The real value isn't just faster code production—it's faster review and testing cycles.
The question is whether this centaur phase will last decades like it did in chess (roughly twenty years) or end sooner. The math cuts both ways. Software engineering is a harder problem than chess and benefits from vastly more funding, which could shorten the centaur age. But more software means more total work, so even if AI handles a bigger percentage, humans might still do more absolute work. The author lists six competing arguments and basically admits nobody knows. Chess is the closest historical parallel and also faced the same automation forces, so assuming a similar timeline—another decade or two—seems reasonable.
For engineers working now, the practical advice is straightforward: don't panic and jump careers. A decade is long enough to make real plans for what comes next or finish out your current career. The mistake is refusing to become a centaur by treating AI as a passing fad—that ship sailed. What matters is figuring out what humans uniquely add. In 2023 it was technical expertise; the author thinks it's shifting toward alignment (making sure AI systems match organizational values), though taste and other factors might matter too. The apocalypse scenario looms on the horizon, but it's been looming for a while. People who bail out early could waste an entire normal career.