Has AI Been Overhyped? Study Shows Minimal Impact On Jobs And Pay
A working paper released by the National Bureau of Economic Research, Denmark, found that AI's use had a negligible effect on hours and pay. On average, employees saved three per cent of their time, while just three to seven per cent of their productivity gains came back to them in the form of higher pay.
"AI chatbots have had no significant impact on earnings or recorded hours in any occupation," economists Anders Humlum and Emilie Vestergaard wrote in the paper.
For the study, 25,000 workers across 7,000 workspaces were analysed. The majority of the employees belonged to occupations (accountants, customer support specialists, financial advisors, HR professionals, software developers and teachers) believed to be susceptible to disruption by AI.
After analysing the data, the researchers said they found no displacement of human workers, nor did they see "transformed productivity and hefty raises for AI-wielding superworkers".
"While adoption has been rapid, with firms now heavily invested in unlocking the technological potential, the economic impacts remain small," the researchers said.
"Modest productivity gains (average time savings of 3 per cent), combined with weak wage pass-through, help explain these limited labour market effects. Our findings challenge narratives of imminent labor market transformation due to Generative AI."
The findings may shock the companies which are actively trying to reduce their headcount in favour of AI. CrowdStrike, the infamous cybersecurity company responsible for the massive global IT outage last year, announced this month that it was slashing five per cent of its workforce and replacing it with AI.
Similarly, language-learning platform Duolingo announced it would "gradually stop using contractors to do work that AI can handle". The company justified its switch in approach, stating that it had taken a similar call in 2012 by betting big on mobile.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Mint
5 hours ago
- Mint
Transformer by Mint: The man shaping India's AI dreams, and continuing chaos at Vodafone
I've known Abhishek Singh, a senior bureaucrat, for some time now. He's been in the Indian tech ecosystem for a while, leading multiple government-backed digitisation initiatives. Now, as chief of the billion-dollar India AI Mission, he faces one of his biggest challenges in a public-service career spanning three decades. The reasons for this are varied. For one, the fact that AI presents a huge opportunity to a long-serving government official shows just how far the technology has come, and how it now affects everyone. More importantly, though, India could potentially gain or lose a lot depending on what we do with AI. Let me take you back a few decades. If you've read the venerable Chip War by Chris Miller (whom I had the pleasure to meet this January), you know that during America's push for leadership in electronic machines at the start of the world's tryst with semiconductors, India missed the bus. This allowed Japan and Taiwan to become global technology leaders despite being societies steeped in tradition. Then came the mobile revolution, and apart from emerging as a big global market, India almost missed the bus there, too. But then the Digital India and Make in India initiatives emerged, digital skills took centre stage, and India is now at a point where tech manufacturing is at least on the ascendancy. To cut a long story short, after having missed out on tectonic global shifts, India a chance to show with AI that it is not just the world's tech back-office and can lead from the front, too. Singh has a plan for this: building a voice-based foundational model that, along with India's government-supported base of thousands of Nvidia GPUs, would become India's next big export to the world after UPI. Here's why he thinks this will work. Speaking of tech's back offices… Jas Bardia, our resident correspondent for India's nearly $300-billion IT services industry, reported last week that there's a war brewing at India's mid-sized tech services firms, which truly believe they can take on the behemoths and win. India's IT services industry had began booming in the early 1990s, turning Tata Consultancy Services, Infosys, Wipro and the likes into the mammoths they are today. During the late 1990s and early 2000s, almost every household around where I grew up had at least one person working at these IT giants. The world, however, as changed considerably since then. Over the past two years companies such as Coforge and Persistent Services have emerged as serious competitors, pitching themselves as specialised firms with a deeper understanding of technology. Where does this leave TCS and its ilk? Will they lose out? Maybe not so soon, but market dynamics are undeniably changing. Also changing is the top job at Vodafone-Idea The beleaguered telecom operator began its India journey as Command Telecom, a telco operated under Kolkata's Usha Martin. In 2000, Hutchison Max acquired Command, leading to the creation of network provider Hutch in 2005. In 2007, Vodafone entered the market and created Vodafone Essar Limited, the entity's longest-standing identity so far. Despite its more than three decades of history, the Vodafone-Idea entity of today is in perilous financial health. Last week the telco appointed erstwhile chief operating officer Abhijit Kishore as CEO for three years as outgoing chief Akshay Moondra's term ended. Now, being a CEO is a dream for anyone in corporate India, but Vi faces a veritable nightmare. After all, it needs to catch up with Airtel and Jio on quality of service while paying off its eye-watering dues and needing $30 billion of capital immediately. Suddenly, Kishore's job doesn't seem like a dream. One thing's clear, though – whichever way this goes, Vodafone-Idea's story will make for a fascinating case study in India's telecom sector for years to come. Mint's telecom correspondent Jatin Grover brings you all the juicy details. Finally, satellites on the frontline Last week, Jatin and I wrote about India's potential revamp of sensitive defence networks in an exclusive report. The full story: over the past two years, the government has been exploring ways for modern satellite internet providers such as Elon Musk's Starlink and Bharti Airtel's OneWeb to offer their services to India's defence forces. The reason is clear: it's now imperative to have secure and blazing-fast internet connectivity even in remote bounary regions. India needs drones, consistent satellite feeds, and a host of other technologies to stay ahead of its enemies. Older satellite connections—which serve only as a backup—aren't up to the task. In other news: the battle for Chrome, and an iPhone 'Air' Last week, Perplexity CEO Aravind Srinivas put in a bid for Google Chrome, saying his company was willing to spend $34.5 billion to buy the world's leading browser. However, he doesn't have that kind of money. You see, Perplexity is only worth about $18 billion. Chrome, on the other hand, is valued more than $50 billion. Then, OpenAI CEO Sam Altman added fuel to the fire, asking, 'Is Google really selling Chrome? If they are, we'd be interested. Why not?" Welcome to Silicon Valley's newest battleground, one that we'll be tracking. We've already reported about Google and OpenAI's silent fight, and how it forced Sergey Brin, a Valley legend, back to the engineering table. Finally, its that time of the year when we expect to see new Google Pixels and Apple iPhones. This year, rumours are that Apple will launch an 'iPhone Air' as part of its range this year. If you've followed Apple, you'd know the 'Air' branding refers to ultra-thin and light devices. The first MacBook Air, in fact, remains one of the most legendary consumer devices to date. Will the iPhone Air live up to this? Here's what we've gathered so far. Transformer by Mint is a weekly newsletter that brings India's most important and interesting technology updates under one umbrella. As the world transforms with every day of innovation, Transformer will keep a tab on the impact that technologies will make in each of our lives. Published every week, the newsletter brings some of India's tech landscape's most insightful coverages until date.
NDTV
5 hours ago
- NDTV
"Unwise For A Human To Copy Paste AI Text": Zoho Chief Sridhar Vembu's Big Warning
Artificial intelligence could "slow us down" if not used properly, Zoho co-founder Sridhar Vembu has said. He also urged individuals against "copy-pasting" AI-generated text as their own. "It is unwise for a human to copy-paste AI text ... hiding the fact that it came from AI," Mr Vembu wrote in a detailed post on X. He shared how he uses AI daily, describing himself as a "moderate to heavy" user with two to three sessions a day. He revealed that he regularly switches between ChatGPT, Claude, Gemini, Perplexity, Grok and also uses Deepseek. Mr Vembu highlighted four main ways AI fits into his workflow. First, he said AI has largely replaced traditional web search, cutting his own search usage by 80 per cent. Second, he uses AI as a "debating partner," often making different models critique each other's responses to deepen his understanding. On business applications, Mr Vembu sees AI as useful in improving product experiences and speeding up customer support. He warned against replacing human agents or passing off AI-generated text as human-written. "It is unwise to let AI replace human agents," he wrote. For coding, Mr Vembu said, while AI can generate code, it requires extensive review for compliance, privacy, and security. "If any programmer submits AI-generated code without doing all this, they are failing at their job," he said, adding that in some cases, AI may even "slow down" development rather than speed it up. He added that at Zoho, programmers are encouraged to use AI mainly for learning and debate. "We continue to run a lot of experiments, and I will revise my opinion if and when facts change on the ground," Mr Vembu said. On our AI use as of August 2025. I use AI chat tools daily, at least 2-3 sessions a day. So I would count myself as a moderate to heavy user. I have the top 5 apps installed in my phone and I use all of them (see item 2). 1. AI helps me learn faster. It is a much better search… — Sridhar Vembu (@svembu) August 14, 2025 Earlier this year, Mr Vembu stepped down as Zoho's CEO to take on the role of chief scientist, focusing on deep tech R&D, particularly in AI, alongside his rural development mission. He said he was preparing to present Zoho's AI strategy to industry analysts in Austin. As part of the leadership transition, co-founder Shailesh Kumar Davey has taken over as group CEO, Tony Thomas now leads Zoho US, Rajesh Ganesan heads ManageEngine, and Mani Vembu oversees
Time of India
6 hours ago
- Time of India
Anthropic's Claude 4 gets feature to cut off abusive user interactions
Synopsis Anthropic framed the move as part of its ongoing research into AI welfare. When Claude ends a conversation, the user can no longer send new messages in that thread.
