Google, Microsoft, and Perplexity Promote Scientific Bias in Search Results

Google added that part of the problem it faces in producing the AI ​​Overview is that, for some specific questions, there is no high-quality information on the web. And there is no doubt that Lynn’s work is not of the highest quality.

“The science behind Lynn’s ‘national IQ’ database is so poor that it’s hard to believe the database is fake,” Sear said. “Lynn never explained how he selected samples from the database; many countries have IQs estimated from absurdly small and unrepresentative samples.”

Sear points out that Lynn’s IQ estimate for Angola is based on information from only 19 people and that of Eritrea is based on a sample of children living in orphanages.

“The problem with the data that Lynn used to create this dataset is bullshit, and it’s bullshit on many fronts,” Rutherford said, pointing out that the Somali population in Lynn’s dataset is based on a single sample of refugees between the ages of eight and eight. 18 were tested in a Kenyan refugee camp. He added that the Botswana scores were based on a single sample of 104 Tswana-speaking high school students aged between seven and 20, who were tested in English.

Critics of the use of national IQ tests to promote the notion of racial superiority point out not only that the quality of the samples collected is poor, but also that the tests themselves are often designed for a Western audience, and are therefore biased before they are even administered.

“There is evidence that Lynn systematically biased the database by favoring samples with low IQs, while excluding those with high IQs, in African countries,” added Sears, a conclusion supported by an earlier 2020 study.

Lynn has published various versions of his national IQ dataset over the decades, the most recent, called “The Intelligence of Nations”, was published in 2019. Over the years, Lynn’s flawed work has been used by food fanatics and racists. groups as evidence to support claims of white supremacy. The data was also converted into a color-coded world map, showing sub-Saharan African countries with a supposedly lower red IQ compared to western nations, which are colored blue.

“This is a data visualization that you see all over Twitter, all over social media, and if you spend a lot of time in racist hangouts on the web, you just see this as a racist argument saying ‘Look at the data. Look at the map,’” Rutherford said.

But the blame, Rutherford believes, lies not only with AI programs, but also with the scientific community that has been uncritically citing Lynn’s work for years.

“Actually, it’s not surprising [that AI systems are quoting it] because Lynn’s work on IQ has been unquestionably accepted in the larger academic community and if you look at the number of times his national IQ data has been cited in academic works, it’s in the hundreds,” Rutherford said. “So the fault is not in AI, the fault is in education.”