Political change vital to democratise AI | Letter
Stephen Cave is right to highlight the ethical issues with artificial intelligence (To save us from a Kafkaesque future, we must democratise AI, 4 January) but we should be wary of focusing on "diversity and inclusion". He argues that including the voices of more women or black people will ensure a more ethical (and efficient) future for AI. But as software developer and activist Nabil Hassein points out: "The liberation of oppressed people can never be achieved by inclusion in systems controlled by a capitalist elite which benefits from the perpetuation of oppressions." Let's not kid ourselves that having female or black CEOs at Facebook would stop its abuses of power.
To democratise AI, we need to take back control of digital infrastructures and build alternatives that serve collective interests. Refugee women in Hamburg and cab drivers in Texas are running cooperative platforms for finding work. Cities such as Barcelona are building identity management systems that give citizens control of their data. Tech and gig economy workers are unionising to demand better rights and hold their bosses politically accountable. These strategies recognise that AI systems relate to wider agendas such as privatisation and deregulation. To alter them we need political change, not just technical fixes focussed on privacy or bias.
Continue reading...