Hindsight · review build · data not yet validated
Although computers still fail the test, confidence is increasing that they will be in a position to pass it within another one or two decades.
Grader A: unfalsifiable
Computers did still fail the Turing test in 2009. But whether 'confidence was increasing' that they would pass within one or two decades has no measure for 2009. Expert opinion was split (the 2002 Kapor-Kurzweil bet framed it as open), and no poll series tracks the trend. A reading about public opinion and a reading about expert opinion give different or ungradable results.
Test: No dated measure of confidence exists for 1999-2009; the Kapor-Kurzweil bet (2002) records disagreement, not a trend.
Grader B: partial
Computers did fail the Turing test in 2009; that part holds. Confidence in passing within 10-20 years (by about 2019-2029) existed in part of the field: a survey of AGI-09 conference experts found most expected human-level AI 'within upcoming decades'. AGI-09 attendees were a self-selected optimistic group, and mainstream AI researchers of the period were more skeptical. No broad measure of rising confidence exists. The claim is partly supported.
Test: Test: a substantial share of experts expect a Turing-test pass within 20 years (by about 2029). AGI-09 expert survey (published 2011): most expected human-level AI within decades. No broader measure for 2009.