Member-only story
Escaping the Sandbox — The Risks of AI From Everyday Disruption to the Loss of Human Control
Artificial intelligence is already changing how we work and live. But the most consequential risks may extend far beyond job losses, misinformation and automation.
Artificial intelligence presents a perplexing challenge. We are adopting technologies that can improve productivity, accelerate scientific discovery and make knowledge more accessible, while simultaneously trying to understand how they might reshape our economies, institutions and relationship with technology.
Some risks are already visible, such as job losses and unemployment, which have been a dominant topic of discussion over the last year or so. Others are emerging gradually however, while a third category concerns future possibilities that remain uncertain but could have extraordinary consequences. Understanding these differences matters greatly. If we treat every AI risk as an imminent catastrophe, we risk losing perspective. If we focus only on problems we can see today, we may fail to prepare for developments that are harder to reverse.
A useful way to make sense of the debate is to distinguish three levels of risk which the progress of artificial intelligence…
