Into the AI Abyss?

From telling jokes to writing code, OpenAI’s GPT-4 seems to be able to do it all – at least enough to displace workers in a wide range of professions than were previously considered relatively safe from automation. The debate over how to regulate generative artificial intelligence is heating up fast, with many prominent voices calling for a moratorium on the development of generative AI technology.

Michael R. Strain, Director of Economic Policy Studies at the American Enterprise Institute, pushes back against this impulse. While “there are times when governments would want to halt a technology’s development,” he writes “this is not one of them,” not least because AI is unlikely to transform the economy quite as fast as its detractors seem to expect. “Regulation should focus on how AI is used,” he argues, “not whether it can continue to develop.”

According to Daron Acemoglu and Simon Johnson of MIT, the way AI is currently being used is deeply problematic: corporate America’s fixation on automation is set to displace and disempower millions of workers. The cost savings may impress investors in the short term, but over time it will erode Americans’ spending power – “the engine of the US economy.”

And yet a “handful of dominant companies” are likely to benefit enormously, notes Diane Coyle of the University of Cambridge, because they will “use deep-learning algorithms” like GPT-4 to “build new services and products….enhance their market power, and erect insurmountable barriers to entry.” This would represent a major missed opportunity to “bring about the productivity growth needed to boost incomes and living standards” – an opportunity that can be seized only if all businesses have access to these “revolutionary tools.”

But Slavoj Žižek, International Director of the Birkbeck Institute for the Humanities at the University of London, argues that generative AI cannot ultimately empower any human – not even “those who develop, own, and control” it – because, by design, it will “need less and less input from human agents.” In the “not-too-distant future,” he concludes, humans could well “become irrelevant, and their lives meaningless.”