MODEL Google/DeepMind

AlphaGo Defeats Lee Sedol

Deep RL at the Game Frontier

DeepMind's Go-playing AI that, in March 2016, defeated 18-time world champion Lee Sedol 4 games to 1. Go had been considered decades away from AI mastery — its branching factor (250+ legal moves per turn vs chess's ~35) had stymied decades of classical AI approaches. AlphaGo combined deep neural networks with Monte Carlo Tree Search to leap past those limits in two years of focused work.

Why it matters

AlphaGo was the first AI system whose capability was undeniable to non-technical audiences worldwide. The cultural moment accelerated AI awareness, funding, and policy attention by perhaps 5 years. Without AlphaGo, the GPT-3 / ChatGPT receptions would have happened against a much less prepared public.

Core Capabilities

Research
Foundational paper or scientific contribution.
Agent Workflows
Built for tool use and autonomous tasks.

Context Window

Context window not disclosed.

Availability

API
Not available
Product / App
Not available
Open Source
Not released
Enterprise

Pricing Model

Research artifact
Not commercially released.
Research

Capability / Performance

Where this model sits relative to the middle 60% of models in the tree. All scores are 0–10 (higher is better).

Lower 20% Upper 80% This model
Lower 20% 20th percentile — 20% of models score below this This model Where the current model lands Upper 80% 80th percentile — only 20% of models score above this Percentile boundaries are computed across every model in the tree that reports the underlying benchmark for each capability.

What it feels like

  • Beat Lee Sedol 4-1 in March 2016 — the moment AI 'solved' Go a decade earlier than experts predicted
  • Watched live by 280 million people; Move 37 in Game 2 became one of the most analysed moves in Go history
  • Combined CNN policy/value nets with Monte Carlo Tree Search — the template later refined into AlphaGo Zero, AlphaZero, MuZero
  • AlphaGo Zero (2017) eliminated human game data — pure self-play beat AlphaGo 100-0 in 40 days
  • Triggered Korea's national AI investment programme and reset global expectations about AI capability
  • Lineage feeds directly into AlphaFold (DeepMind's biology breakthrough) via shared search-and-evaluate machinery

Best use cases

  • Inspired the entire deep-RL research wave 2016-2020 (DeepMind, OpenAI, others)
  • AlphaFold (same lab, similar compute substrate) (DeepMind)
  • MuZero, AlphaStar, AlphaCode (DeepMind lineage) (DeepMind)

Tools to try

Not ideal for

  • Tasks far outside the modalities listed in this model's spec
  • Workflows where a more recent successor in the same family scores higher

Model Evolution

View full evolution tree →

Silver, D. · Huang, A. · Maddison, C. J. · Guez, A. · et al.