superhuman coder
The first takeoff milestone: an AI that can do coding tasks better than the best humans, providing roughly a 5x algorithmic progress speedup.
As defined in AI 2027: month-by-month model of intelligence explosion — Scott Alexander & Daniel Kokotajlo
Where it comes up · 3
- What Will AI Look Like in 2027? | Interview3 claims
- AI 2027: month-by-month model of intelligence explosion — Scott Alexander & Daniel Kokotajlo1 claim
- AI 2027: month-by-month model of intelligence explosion — Scott Alexander & Daniel Kokotajlo1 claim
Most important claims · 5 of 5Search all claims for “superhuman coder” →
- The takeoff is modeled as three milestones — superhuman coder (~5x algorithmic speedup), superhuman AI researcher with the whole research stack automated (~25x speedup), and superintelligent AI researcher (~hundreds to 1000x overall speedup) — with each milestone's speedup feeding into reaching the next.AI 2027: month-by-month model of intelligence explosion — Scott Alexander & Daniel Kokotajlo
- Once superhuman coders exist, they would be deployed to automate the full AI research process within roughly six months (halfway through 2027 in the scenario), achieving 25x faster algorithmic progress while compute scaling remains flatWhat Will AI Look Like in 2027? | Interview
- One key difference between the AI 2027 scenario's takeoff and previous technological revolutions (Industrial Revolution, Agricultural Revolution, Cambrian Explosion) is that previous transitions were smooth hyperbolic curves with no hard discontinuities, whereas AI takeoff involves discrete capability thresholds (superhuman coder → automated research → superintelligence) that create discontinuous speedups.AI 2027: month-by-month model of intelligence explosion — Scott Alexander & Daniel Kokotajlo
- The AI 2027 scenario breaks down AI capabilities development into discrete milestones (superhuman coders, superhuman AI researchers, broad superintelligence) rather than a single AGI threshold, providing more granular prediction pointsWhat Will AI Look Like in 2027? | Interview
- Cocatello agrees with Aur's critique and argues that superhuman coders must be augmented with reinforcement learning for long-horizon agentic tasks, representing a meaningful departure from current language model architectures.What Will AI Look Like in 2027? | Interview