Turing's 1948 Imitation Game: A Human-Approximates-Machine Reading
An academic article recently published on arXiv (2608.05558) offers a fresh perspective on Alan Turing's 1948 document 'Intelligent Machinery,' framing it as a key reference for his subsequent imitation games through a new 'human-approximates-machine' lens. The paper first highlights and merges core design elements of the 1948 chess-based imitation game, such as the potential for intelligent machines to err, the omission of irrelevant physical traits, the significance of the human judge, and Turing's assertion that intellectual activity largely involves search. Additionally, it posits that limiting the human participant to a subpar chess player enhances the importance of intellectual search, making human behavior more akin to that of machines. This interpretation indicates that the imitation game framework can also explore when human intelligence resembles machine-like behavior in specific tasks. The paper, classified as a cross-type submission, adds to the discourse on artificial intelligence and the philosophy of mind.
Key facts
- Paper examines Turing's 1948 report 'Intelligent Machinery' as a conceptual source for later imitation games.
- Identifies design concepts: intelligent machines may make mistakes, exclusion of irrelevant physical features, role of human judge, intellectual activity as search.
- Argues restricting human contestant to a poor chess player increases intellectual search and makes human behavior more comparable to machine behavior.
- Presents the 1948 game as a human-approximates-machine game.
- Suggests imitation game framework can examine when human intelligence becomes machine-like.
- Paper is on arXiv with identifier 2608.05558v1.
- Announce type: cross.
- Published on arXiv (URL: https://arxiv.org/abs/2608.05558).
Entities
Artists
- Alan Turing
Institutions
- arXiv