As automation advances, many researchers have predicted widespread job losses, with estimates ranging from 7 percent of jobs in the next decade to 850,000 by 2030. This has fueled calls for a universal basic income (UBI) to cushion the impact. But a Duke University professor argues that current AI is not capable of replacing most workers, making UBI an unnecessary response at this time.
Vincent Conitzer, who holds professorships in computer science, economics, and philosophy at Duke, said in an MIT article that while AI has made impressive strides, it still falls short in areas like understanding social norms and picking up on subtle social cues. He highlighted language comprehension as a key weakness, noting that computers struggle to answer even simple questions because they lack a broad understanding of the world and common sense.
Conitzer also pointed out that AI systems cannot engage in true abstraction or examine their own reasoning, which limits their creativity. He cited Google DeepMind's AlphaGo and DeepDream as examples, saying their ability to think outside the box is not the kind of creativity that offers a genuinely new perspective. These limitations, he argues, make AI ill-suited for the messy realities of most workplaces.
Why UBI Is Not the Answer Yet
Given these constraints, Conitzer does not see the need for a basic income at present. He acknowledges that AI could eventually eliminate some jobs, particularly those with well-defined tasks. However, he believes many roles—such as therapists, counselors, and kindergarten teachers—are safe because they require a general understanding of the world and people, something AI cannot replicate.
Conitzer admits that AI could progress faster than he expects, but he does not foresee it surpassing human capabilities in the short term. He dismissed the idea that recent AI progress will prevent most people from meaningfully contributing to society as nonsense. Still, he advises monitoring developments and preparing for potential shocks.
Preparing for an Uncertain Future
Conitzer suggests that society may need to adapt, including making it easier for displaced workers to retrain and increasing public spending on carefully selected infrastructure projects to counterbalance job losses. He also emphasizes the importance of building resilience to unexpected AI advances.
While such preparation is prudent, the full impact of AI on the workforce remains uncertain. Until then, Conitzer's advice is simple: don't quit your day job.
