IBM Watson defeats human champions on Jeopardy!

事件摘要

On February 16, 2011, IBM's Watson supercomputer defeated all-time Jeopardy! champions Ken Jennings and Brad Rutter across three episodes. Watson used natural language processing and information retrieval over 200 million pages of content, demonstrating that machines could handle the ambiguity of natural language at human-comparable levels in a constrained domain.

影响评估

• Capability Leap +1 · Medium-term

  Demonstrated that NLP + information retrieval could parse natural language ambiguity in a competitive real-time setting. However, the approach (hand-engineered features, massive compute, no real learning) proved to be a dead end compared to deep learning alternatives that emerged shortly after.

  Affected Groups: NLP researchers, IBM, computer scientists

• Paradigm Shift +1 · Short-term

  Publicly demonstrated that machines could compete with humans on general knowledge quizzes involving natural language understanding. Shifted public perception from 'AI can play chess' to 'AI can answer any question.'

  Affected Groups: general public, media, business leaders

共识度与来源

• 1
  "Watson, Jeopardy! champion" — IBM History

  URL: https://www.ibm.com/history/watson-jeopardy

  Watson defeated the two most celebrated champions in Jeopardy! history—Ken Jennings and Brad Rutter—in a three-day event.

  Reference Evidence Citation logged Live source
• 2
  "Looking back at Watson's 2011 'Jeopardy!' win" — Axios (2021)

  URL: https://www.axios.com/2021/02/13/ibm-watson-jeopardy-win-language-processing

  IBM has spent years trying to make its Watson technology a business success. The results have been mixed.

  News Report Citation logged Live source