Human vs. Computer Go: Review and Prospect [Discussion Forum]
Level 5 - mechanism / opinion, no new human data
By design analogy: narrative review, discussion forum, and qualitative expert commentary with no clinical or experimental trial data.
OpenAlex W2414015145 · doi:10.1109/mci.2016.2572559
What was done
The authors reviewed the historical milestones in computer Go development from 1998 to 2016, surveyed competitive Go programs from IEEE Computational Intelligence Society competitions, described the computational intelligence techniques used in AlphaGo, and compiled commentaries from three high-level professional Go players on the five AlphaGo versus Lee Sedol games in March 2016.
What was found
The abstract reports no quantitative performance metrics, game outcome counts, or statistical analyses. It provides a qualitative narrative of computer Go milestones leading to AlphaGo's match against Lee Sedol and summarizes professional Go players' commentaries.
Why it matters
The paper captures the technical developments in computational intelligence leading to AlphaGo's milestone performance and contextualizes machine capability through the perspective of elite human players.
Limits
This is an expert commentary and narrative review rather than a systematic review or controlled experimental trial. The evaluations are qualitative, based on a single five-game match, and provide no formal statistical metrics in the abstract.
Cited by
- supports The AI program AlphaGo defeated the world champion in the board game Go.