The Turing test has sometimes been called Alan Turing's worst idea.
Unresolvable The public record could not settle it.
Factual: Statements about the present or the past, checked against the record: did they hold up?
What would count as true or false
TRUE: at least two independent published sources dated before or during 2023 (articles, books, talks by scholars or writers) describe the Turing test as Alan Turing's 'worst idea' or use near-identical wording. MOSTLY TRUE: exactly one identifiable published source uses the phrase, or sources use close but somewhat different wording (e.g. 'one of Turing's worst ideas', 'Turing's biggest mistake'). MOSTLY FALSE: the test has been criticized harshly, but nobody has called it Turing's worst idea or anything close to it. FALSE: no such description exists before 2023, and the evidence shows the phrase began with Pinker himself or was attributed to a source that does not contain it.
Set before the research began.
Evidence (7)
-
E1 Podcast: The case for global optimism with Steven Pinker GZERO Media · 23 September 2023 Own record
“Some people called it Turing's worst idea because it all depends on, how easy is to fool human? And the answer is pretty easy.”
-
E2 Turing Test Considered Harmful IJCAI (International Joint Conferences on Artificial Intelligence) proceedings; authors Patrick Hayes and Kenneth Ford · 1995
“The Turing Test had a historical role in getting AI started, but it is now a burden to the field, damaging its public reputation and its own intellectual coherence.”
-
E3 Podcast: The case for global optimism with Steven Pinker GZERO Media · 23 September 2023 Own record
“Some people called it Turing's worst idea because it all depends on, how easy is to fool human? And the answer is pretty easy.”
-
E4 Turing Test Considered Harmful IJCAI (Patrick Hayes and Kenneth Ford, Beckman Institute and Institute for Human & Machine Cognition) · 1995
“The Turing Test has been with AI since its inception, and has always partly defined the field.”
-
E5 Human Intelligence, Artificial Intelligence, and Homo sapiens ... Harvard Data Science Review (MIT Press) · 27 October 2023
“Pinker reminded me that the Turing test has sometimes been considered Alan Turing’s worst idea because “It’s a test of how easily you can fool people, and the answer is ‘easily’” (Pinker et al, 2023).”
-
E6 GZERO WORLD with Ian Bremmer | Is the World Actually Getting...Better? | Season 2023 PBS (GZERO World with Ian Bremmer, THIRTEEN PBS) · 22 September 2023 Own record
“Some people called it Turing's worst idea because it just all depends on how easy it's to fool a human.”
-
E7 Goodbye, Turing Test; Bring on the Turing Decathlon - Nautilus Nautilus · 6 February 2015
“but every one of us in the room thinks it’s an illegitimate test, and that it’s a poor goal for artificial intelligence.”
Reasoning
The 2023 GZERO and PBS items (E1, E3, E6) contain only Pinker's own statement that "some people" called the Turing test Turing's worst idea, and they name no earlier source. The Harvard Data Science Review editorial (E5) uses nearly the same wording, but it credits the remark to Pinker's 2023 interview rather than to any earlier writer. Hayes and Ford (E2, E4) and the Nautilus article (E7) criticise the test harshly, but neither calls it Turing's worst idea. Because no independent source from before 2023 uses the phrase, and nothing shows that the phrase began with Pinker, the evidence doesn't settle the claim.
Cite this claim
Steven Pinker, “The Turing test has sometimes been called Alan Turing's worst idea.” (said 2023). Verdict: unresolvable. Reliability API, claim c_6044737470ce, method 0.1.0+28da144b, dataset release v1.1.2. https://epistacklabs.com/claims/c_6044737470ce
This claim is in dataset release v1.1.2, with the same verdict.
Reply to this claim
More factual claims by Steven Pinker
-
Mostly true
“A major fact about language, known for decades, is that people don’t blurt out what they mean”
-
Mostly true
“Computers have always been better at some of the things that humans do, otherwise we never would have adopted them”
-
Resolved true
“then it is very much unlike AlphaGo, with this closed universe governed by fixed rules.”
AlphaGo operated in a closed universe governed by fixed rules, unlike the open-loop real world.
-
Mostly true
“The capabilities of LLMs trained on gargantuan datasets are more general than I (and almost everyone else) would have guessed at the time.”
-
Mostly true
“There are some features in large language models that actually implicitly build in some of the insight of linguistics.”
Verdicts are made automatically; how verdicts are made.