Why We Keep Moving the Goalposts on Artificial Intelligence

The XandY Poll | U.S. Adult Population

Abel Gustafson, Matthew Goldberg, & Kyle Gustafson
December 2024

In the rapidly evolving field of artificial intelligence, there's a curious phenomenon known as "The AI Effect" or "Tesler's Theorem": people consider some technology to be AI right up until it becomes commonplace, at which point people stop considering it AI. As computer scientist Larry Tesler famously observed, "Intelligence is whatever machines haven’t done yet." 

Our data from the XandY Poll illustrates this effect in action. Despite the sophisticated algorithms powering many of the technologies all around us, Americans are surprisingly reluctant to classify them as AI. Navigation apps that calculate optimal routes and adjust in real-time? Only 31% of Americans consider that artificial intelligence. Language translation tools? Just 38% view them as AI. Even spelling and grammar checkers, which employ complex natural language processing, are seen as AI by just 39% of respondents.

The pattern becomes even more interesting when examining technologies more commonly associated with AI in popular media. While chatbots and virtual assistants do get majority recognition as AI (61% and 60% respectively), these margins are surprisingly slim. Even humanoid robots that can engage in conversation – perhaps the most iconic image of AI in popular culture – are considered AI by only 58% of respondents.

This contemporary data demonstrates Tesler's observation about how we conceptualize AI. Technologies that were once considered remarkable achievements in artificial intelligence, like language translation, adaptive navigation, or game-playing computers, are now so integrated into our daily lives that many people no longer view them as AI at all. Instead, the label of "real AI" gets pushed forward to newer, less familiar technologies.

More than 20 years ago, Rodney Brooks, the former director of MIT’s Artificial Intelligence Laboratory said "Every time we figure out a piece of [AI], it stops being magical; we say, 'Oh, that's just a computation... We used to joke that AI means 'almost implemented.'" 

The implications of this perspective shift are significant. This tendency to continuously move the goalposts of what constitutes "real AI" can make it difficult to have meaningful discussions about artificial intelligence and its impacts. When successful AI applications are routinely reclassified as just another computer program, we risk underestimating both the progress we've made as well as the real capabilities of AI systems.

About The XandY Poll

SURVEY METHODOLOGY

The design, data collection, analysis, and reporting of this national survey were performed by XandY, an independent research and strategy firm. Exhaustive details of the scientific methodology of The XandY Poll can be found here. The following briefly summarizes key points of interest.

The survey responses were fielded from November 18 – November 21, 2023 using online recruitment methods to sample adult residents of the United States (N = 1,527). This survey used a nested quota sampling strategy to match U.S. Census proportions of age, income, race and ethnicity, gender, and political party affiliation. To further ensure the insights reported from these data closely resemble the U.S. population, the sample was weighted to match U.S. Census benchmarks.

MARGIN OF ERROR

Proportion statistics regarding the full national sample have an average margin of error of +/- 2 percentage points at the 95% confidence level. The margin of error in subgroups is determined by the subgroup size.

CITATION

This paper and the insights it reports may be cited as:

Gustafson, A., Goldberg, M. H., & Gustafson, K. (2024). Why We Keep Moving the Goalposts on Artificial Intelligence. XandY. New Haven, CT. Retrieved from: https://www.xandyanalytics.com/moving-the-goalposts-on-ai