Superhuman coder emerges

What AI 2027 Predicted

The scenario predicts the emergence of a “superhuman coder” — an AI system that surpasses the best human programmers on essentially any coding task, both in quality and speed. This is a key milestone in the path toward broader superintelligence.

How We Track This

We monitor:

• SWE-bench Verified and SWE-bench Pro scores
• Terminal-Bench results
• Real-world coding competitions (Codeforces, etc.) — AI vs human rankings
• Enterprise reports on code quality from AI vs human developers
• Novel system-level projects completed entirely by AI

Current Evidence

Coding AI is advancing rapidly but “superhuman” remains far. Claude Opus 4.6 (Thinking) leads SWE-bench Verified at 79.2%, GPT-5.4 at 77.2% (vals.ai). But on the harder SWE-bench Pro (real-world complexity), best scores are only 23.3% (GPT-5) and 23.1% (Claude Opus 4.1), per Scale Labs. Terminal-Bench 2.0: GPT 5.3 Codex at 65%, Opus 4.6 at 63%. 16 Claude Opus 4.6 agents wrote a C compiler from scratch. Claude Code went from zero to #1 coding tool in 8 months. But the gap between “very useful coding assistant” and “superhuman coder” (any task, faster and cheaper than best human) remains large.

Sources:

• SWE-bench Verified Results — Vals.ai
• SWE-Bench Pro Leaderboard — Scale Labs
• Inside OpenAI’s Race to Catch Up to Claude Code — WIRED
• Grading AI 2027’s 2025 Predictions — AI Futures Project
• Anthropic: Introducing Claude Opus 4.7

Counterevidence & Limitations

• SWE-bench Pro results (~23%) show that real-world coding is far harder than benchmarks suggest
• “Superhuman” is a high bar — surpassing the best humans on any task is qualitatively different from being a useful assistant
• Current tools require significant human guidance for complex projects
• The March 2027 predicted date may be too aggressive by 6–18 months

What Would Change Our Assessment

• Upgrade to “emerging”: SWE-bench Pro scores above 50%; AI consistently winning coding competitions
• Upgrade to “on-track”: SWE-bench Pro above 70%; credible reports of AI completing complex projects without human guidance
• Maintain at “not-yet-testable”: Prediction date hasn’t arrived yet

Update History

Date	Update
2026-04-27	Anthropic released Claude Opus 4.7, highlighting better performance on difficult and long-running software engineering tasks, including customer reports of hours-long autonomy, self-verification, and stronger async coding workflows (Anthropic). This modestly strengthens the evidence that coding models are advancing toward more autonomous engineering work, but does not establish the cheap superhuman-coder threshold.
2026-04-02	AI Futures Project Q1 update: Kokotajlo shifted Automated Coder median from late 2029 to mid 2028 (1.5 years sooner). Revised AC time horizon requirement from 3 years to 1 year, citing Opus 4.6 impressiveness. Eli Lifland’s AC median also pulled forward 1.5 years. The scenario’s own authors are converging faster toward the original timeline. Source: LessWrong
2026-03	Prediction timeframe not yet reached (March 2027). AI coding capabilities advancing rapidly — SWE-bench scores improving, autonomous coding agents shipping — but superhuman performance across all coding tasks remains distant.
2026-01	Kokotajlo personal median for full coding automation: December 2030.
2025-12	AI Futures Project places median for “Superhuman Coder” at December 2031 — vs AI 2027 scenario’s January 2027.
2025-11	Claude Opus 4.5 reportedly outperforms every human candidate on Anthropic internal engineering assessments. Gemini 3 scores 37.4% on Humanity’s Last Exam (world record). Gap narrowing visibly but “superhuman coder” milestone remains contested.
2025-09	Gemini 2.5 Deep Think achieves gold-medal performance at 2025 ICPC World Finals (Sep 17), solving 10/12 problems including one no human team solved. Strongest “superhuman” coding signal to date, though competitive algorithmic programming differs from real-world software engineering.