A Nobel Laureate offers a more modest forecast on AI’s Wildest Promises
Key points
• Economist Daron Acemoglu says AI will automate far fewer tasks than current hype suggests
• Predicts only a 5 percent automation impact and roughly 1 percent boost to global GDP over ten years
• Urges leaders to focus on human centered innovation instead of cost cutting fueled by algorithms
The story
There is no shortage of bold predictions about how artificial intelligence will reinvent the global economy. But Nobel laureate and MIT economist Daron Acemoglu is urging leaders to take a breath.
In a new MIT Sloan Management Review video, he lays out a far more grounded view of what AI will actually deliver in the decade ahead. Based on years of research into technology and labor markets, Acemoglu argues that the economic upside of AI has been overstated. His estimate is that only about 5 percent of all tasks are realistically automatable with current techniques. That figure is far from the sweeping narratives that suggest most white collar work is on the brink of replacement.
He goes further, predicting that AI may add roughly 1 percent to global GDP over the next ten years. It is a meaningful contribution, but nowhere near the internet scale disruption many expect. In his view, AI simply is not yet producing the kind of broad, general purpose applications that transformed industries during the early digital era.
A big part of the challenge is tacit knowledge. Many real world tasks depend on context, interpersonal understanding and subtle judgment, which do not translate easily into models. Acemoglu argues that this complexity is why AI still struggles in many professional environments and why human oversight remains essential.
Instead of racing toward algorithmic cost cutting, he encourages companies to look for ways AI can support workers and unlock new services. The real opportunity, he says, comes from human AI collaboration, not from treating AI as a shortcut to smaller payrolls.
For leaders trying to separate signal from noise, his message is simple: stay curious, stay cautious and do not mistake hype for strategy. Watch the video (15 mins)
Timestamps
0:00 – Introduction: AI’s economic impact predictions
0:34 – Acemoglu’s 5% automation prediction
1:15 – Why Acemoglu’s estimates differ from others
2:27 – Why AI applications aren’t yet transformative
3:12 – Comparing AI’s impact with the internet’s
3:42 – Which tasks AI can and cannot automate
4:47 – How Acemoglu arrived at the 5% prediction
5:54 – The challenge of tacit knowledge in occupations
7:16 – The complexity of real-world tasks
8:35 – AI’s effect on jobs in the next decade
9:29 – A more pro-human approach to AI
9:56 – AI’s potential to create new services
10:53 – Advice for business leaders: beyond the hype
12:47 – Avoiding blind AI investments
14:11 – Working with employees to identify AI value
Source: MIT Sloan Management Review, May 28, 2025
