Share
Reported by Bekki Parrish NERC-funded Policy Intern (May - July 2019) and Kate McNeil, CSaP Communications Coordinator.
Stephen Cave contemplates “Intelligence as Ideology”
At a plenary session hosted by CSaP, Dr Stephen Cave, Executive Director of the Leverhulme Centre for the Future of Intelligence implored the audience to use the concept of intelligence carefully, and to consider whether its usage is appropriate or whether it is being used to reinforce privilege or legitimise oppression.
The discussion, conducted with the potential future implications of artificial intelligence for humanity in mind, examined the idea that some also consider new forms of artificial intelligence (AI), to pose considerable threat to humanity. These ‘doomsday’ perceptions of AI, grounded in reactions to the notion that intelligence and capabilities of the computer outstrips that of humankind, stem in part in the way we understand, and have historically understood, what intelligence is, Dr. Cave argued.
Dr. Cave described intelligence as being viewed by society as a general, measurable capacity (i.e. IQ tests), which are highly valued for determining individual potential and life trajectories. He traces the roots of this mindset back to the high value ascribed to ‘thinking’ by Plato and Aristotle; and Descartes’ equation of intelligence with identity vis a vis “I think therefore I am”. Dr. Cave also highlighted that the history of “intelligence” as a concept has included a problematic entanglement of ideas of intellect with eugenics and racial science built on notions of colonialism, white supremacy, a belief that intellect was racially inherited, and the belief that poverty and intellect were interrelated. Early variants of the Stanford-Binet intelligence test (IQ test) and SATs perpetuated these concepts through culturally biased approaches to measuring intellect, while extreme right-wing groups continue to perpetuate these ideas even though they have long since been discredited by the scientific community.
This history needs to be kept in mind as society goes forward in developing AI, and as policymakers examine opportunities for its use. Cultural bias is a critical challenge facing emerging AI technology, because of their reliance upon data sources which perpetuate historic biases. This presents challenges in the use of AI technology in areas ranging from predictive policing algorithms to CV sorting algorithms. According to Dr. Cave, "the real impact of this technology is disproportionately on those are already marginalized." This will be compounded by this liklihood that public services will be increasingly automated.
Dr. Cave implored those working in this space to consider what the increased use of AI will mean for the communities who are most in contact with AI-driven services, and to examine the impact that it will have upon marginalized populations, including those who have already suffered at the hands of racial profiling, or are of gender, race, or class that has faced historical discrimination or oppression.
You can listen to the complete session here.
Photo credit: Hartwig HKD, 2010