GenAI's Impact on U.S. Productivity

Improving productivity is a crucial aspect of enhancing company efficiency. Generative AI (GenAI) tools like ChatGPT and tailored LLM models hold great promise in achieving this goal. A recent blog post by Intuit's data team explores their study investigating the impact of these tools on data analyst productivity. The study recruited several internal analysts from different business units, spanning various tenures and levels of analytics experience, to ensure diverse participation. Half of the analysts were given access to an internal GenAI tool and tasked with completing representative work assignments within an hour. The study carefully balanced tasks involving both familiar and unfamiliar domains to account for domain expertise. The results revealed a significant productivity increase among GenAI tool users, with SQL tasks being completed 2.2 times faster, or a 55% reduction in time, compared to the control group. Interestingly, the study found that junior analysts experienced the most substantial productivity gains, as well as those tasks involved handling unfamiliar data. This study sheds light on effective approaches to measuring productivity enhancements in the data analyst domain. Despite potential issues with hallucination and accuracy in GenAI tools, their integration with proper user experience interaction proves highly beneficial for productivity enhancement. As more industrial-customized LLM models emerge, they may herald a forthcoming trend in elevating productivity in the analytical domain. #data #analytics #llm #generativeai #productivity #experiment – – –  Check out the "Snacks Weekly on Data Science" podcast and subscribe, where I explain in more detail the concepts discussed in this and future posts:    -- Apple Podcast: https://lnkd.in/gj6aPBBY    -- Spotify: https://lnkd.in/gKgaMvbh https://lnkd.in/gAxNBbCg

🚀 How past tech disruptions can help inform the economic impact of AI via EY 💻 In recent years, no technology has created more excitement than generative artificial intelligence (#GenAI), but that excitement has been tempered by uncertainty and concerns among executives, policymakers and other stakeholders. 🤖 GenAI systems are so complex and developing so rapidly that it is difficult to predict how they will impact organizations, economies, and societies. In this first article of the series, my EY-Parthenon colleague Lydia Boussour uses history as a guide to shed light on the potential future impact of GenAI and the economic opportunities and challenges that it may bring. 📈 Three key lessons from past episodes of rapid technological change can help inform how AI may affect the #economy: 1️⃣ Significant productivity boost: GenAI will likely lead to a significant acceleration in productivity growth and raise living standards like prior general-purpose technologies. By examining the 1990s IT-driven acceleration in productivity growth, we estimate that GenAI could lift productivity growth by 20% to 50% in the coming decade. However, it will likely fall short of the doubling or tripling of productivity growth resulting from the Industrial Revolution or adoption of electricity. 2️⃣ Potentially delayed impact: The productivity boost from GenAI will likely occur with a lag, but the faster speed of technological diffusion and adoption could mean that the boost to economic activity is felt in the next three to five years versus multiple decades for the steam engine and 10 years for the computer age. 3️⃣ Nuanced job reshuffling: AI technologies are poised to cause significant labor market disruptions by automating some tasks and displacing workers, but it will also create new types of jobs and functions within roles across many sectors of the economy that will help offset AI-related job losses. Read the full article here: https://lnkd.in/d9ae9HRi

The exact scale of AI's eventual impact on productivity is uncertain, and economists have wide-ranging estimates. What we do know is that the impact will depend upon how quickly AI spreads, how rapidly the models advance, how people and organizations choose to use it, and the policies that shape its deployment. With 2.5 billion messages sent to ChatGPT globally each day—including 330 million in the U.S.—OpenAI has a unique lens into how people are actually using AI today. In this new economic analysis, OpenAI explores early signals from ChatGPT usage data and what they might mean for productivity, economic growth, and the future of work. Key insights: - Top personal productivity use cases include Learning & Upskilling (20%), Writing & Communication (18%), and Programming, Data Science & Math (7%). - AI natives are emerging: One quarter (24%) of US users are between the ages of 18 and 24, and one third (32%) are between ages 25 and 34. This means that many students and workers in the early stages of their careers are becoming AI natives who will bring this expertise to their careers for years to come. - Small business adoption: ~40% of small U.S. businesses currently use AI; many rank it among the top tools for future success. - Early economic benefits are strongest in sectors with language- and pattern-heavy work: Think legal services, education, government, customer support, consulting, and marketing—where tasks like summarizing documents, generating content, and answering questions dominate. Bottom line: We're seeing real data on how AI is already transforming work across industries. The question isn't whether AI will boost productivity—it's how quickly these gains will scale. How has AI changed your daily work? What productivity gains have you experienced? #AI #Productivity #FutureOfWork #DigitalTransformation

GenAI has indeed put CIOs on the hot seat 🥵 CEOs and their executive peers are looking at CIOs to drive the generative AI strategy for the business. In my interactions with CIOs over the last 6 months, it’s apparent that the role of the CIO has become more relevant than ever. One value stream that shows up consistently is employee service. Why is this a lucrative area for applying Generative AI? 1. Reduce the high spend on employee service desks 2. Create employee productivity surplus 3. Repetitive nature of tasks lends itself to GenAI application 4. Safe area to start using GenAI 5. Excellent opportunity to learn at scale As a result, it's not surprising to see CIOs setup audacious goals like “zero service desk” as they look at the next 2-3 years. They are also clear eyed about the requirements of a generative AI service desk. They don’t want to “sprinkle” a copilot and hope for transformative results. Instead, here is what they are thinking of 1. Search across variety of documents and manuals across dozens of enterprise systems, and summarize answers (eg troubleshooting VPN, benefits policies) 2. Perform actions on behalf of employees to resolve their issues (eg provision applications, change permissions, book time off) 3. Provide real time business data to employees securely and responsibly (eg PO / invoice data, or time off data) 4. Enable their engineers to extend the copilot for new workflows and automations (eg enable sellers to submit requests for deal discounts) 5. Empower tech writers, service owners, HRBPs to create new knowledge / content faster 6. Provide insight to services leaders to help them identify new areas of opportunity for further automation 7. Have up to date success plans that continuously transform their service desk 8. Ensure data and information security while achieving this transformation The next few years will be transformative for employee service. It is emerging as the only consensus value stream for all CIOs to target with GenAI. https://lnkd.in/gg6T2Vky

The numbers have started coming in about the impact of Generative AI on work productivity and quality - 40% increase in productivity and 18% better quality for professional writing tasks. Here is the abstract from a paper published in the prestigious journal Science by Shakked Noy and Whitney Zhang of MIT. We examined the productivity effects of a generative artificial intelligence (AI) technology, the assistive chatbot ChatGPT, in the context of midlevel professional writing tasks. In a preregistered online experiment, we assigned occupation-specific, incentivized writing tasks to 453 college-educated professionals and randomly exposed half of them to ChatGPT. Our results show that ChatGPT substantially raised productivity: The average time taken decreased by 40% and output quality rose by 18%. Inequality between workers decreased, and concern and excitement about AI temporarily rose. Workers exposed to ChatGPT during the experiment were 2 times as likely to report using it in their real job 2 weeks after the experiment and 1.6 times as likely 2 months after the experiment. #AI #artificialintelligence #chatgpt https://lnkd.in/gQiQ2Dek

Study 2 on GenAI productivity impact. This one is a paper by Erik Brynjolfsson, Danielle Li, and Lindsey Raymond (authors from Stanford and MIT), and it’s focused on contact center productivity gains using GenAI tools. Apologies it’s so late, I meant to post on it a month ago! You can probably transfer lessons from this study to IT contact centers, Retail contact centers, Financials contact centers, etc. So it’s applicable to multiple industry horizontals (CIO offices, Customer Support, etc). It’s also fairly representative as it covered the work of about 5000 agents. This is in contrast with some GenAI studies that have very small representation (50-100 subjects with control and experiment group). Let me skip to the chase (if you don’t know the expression, it comes from a Steve McQueen movie!). The conclusion was that contact center staff gained a 14% productivity bump on average, but this result is masking an uneven distribution. Low skilled staff (with less than 2 months experience on this job) gained a 35% jump while very experienced staff gained a negligible improvement. This is a huge impact given that in this industry, the authors estimate, there is a substantial turnover rate in staffing in contact centers, so there’s a large number of new on-the-job staff. The theory behind this difference in contact center agent segments is that inexperienced staff are being augmented with what to say (based on prior successful contact resolutions) while experienced staff are already very good at knowing what to say and need no help. One of the surprising results for me was the substantial increase in sentiment for customers. The authors studied the sentiment in the chats from customers and used sentiment measuring software to score it. Customer sentiment scored half a standard deviation higher when agents started using the GenAI support compared to when agents were not using it (before introduction). In addition there was a substantial decrease in manager escalation. (surprisingly, NPS scores didn’t budge). So the takeaways are that the productivity impact is substantial (14% across the contact center), and even bigger if one focuses this on low productivity novice users (35%). One needs to be thoughtful of what segment to dedicate these tools towards (less experienced staff). And that one shouldn’t measure purely time savings, but also count customer sentiment, manager escalation, and other such measures. For enterprises where contact / call center support is a substantial part of the business (b2c vs b2b) the benefits from introducing GenAI tools appears to be substantial. What do you think about this study? Interesting? Was it what you expected? #genai #aistrategy https://lnkd.in/gTGCUp8t

"This “use-case sprawl”...can be divided into three big categories: window-dressing, tools for workers with low to middling skills, and those for a firm’s most valuable employees. Of these, window-dressing is by far the most common." The Economist is breaking down the claims of GenAI in enterprise. One year into the GenAI revolution and it's time to start taking stock of what is real and what was just clever noise. 𝐔𝐬𝐞 𝐜𝐚𝐬𝐞 𝐬𝐩𝐫𝐚𝐰𝐥 𝐢𝐬 𝐑𝐄𝐀𝐋 Everyone from JPMorgan (300!), Capgemini (500!) to Bayer (I see your 500, and raise you 700!) claimed hundreds of use cases, but how many still matter? Data is coming out of everyone from Air Canada to Instacart, pulling back from some of these initial experiments. Does this mean GenAI is useless? No! But there was very much of a "shiny object looking for a problem to solve" mentality last year. We are seeing enterprises start to refocus on AI (GenAI and otherwise) with a smaller list of use cases that matter to the bottom line. 𝐑𝐞𝐬𝐮𝐥𝐭𝐬 𝐚𝐫𝐞 𝐑𝐄𝐀𝐋, 𝐛𝐮𝐭 𝐧𝐨𝐭 𝐚𝐬 𝐫𝐞𝐚𝐥𝐥𝐲 𝐛𝐢𝐠 𝐚𝐬 𝐨𝐧𝐜𝐞 𝐜𝐥𝐚𝐢𝐦𝐞𝐝 The much-hyped numbers of 70% of employees' tasks being automated by AI isn't really tracking. In November, an enterprise survey "found that just 12% of corporations believed that generative AI had replaced human labour or would replace it within 12 months." Instead of replacing people, many companies are excited at 10 - 20% improvements in their existing teams' productivity and process, and sometimes noting that bottlenecks are shifting to another part of the organization. These incremental improvements are really happening - especially in writing, coding, document processing, and data extraction - with additional less tangible "...big improvements in establishing an internal “data and analytics culture," which will undoubtedly pay dividends in years to come. Enterprises should focus on realizing 10-20% gains and succeed rather than chase an unrealistic 70% and fail. There is a lot of value in pursuing and AI strategy to gain operational efficiency, but REAL results will take measured planning and execution - not just throw some AI at that thinking. https://lnkd.in/gDWeRzHr 

I didn’t expect one of the more interesting & insightful papers on GenAI in the past few months to come from the Federal Reserve, but here we are. Makes a case similar to the one Jeffrey Ding has made so compellingly in his works on the diffusion of AI. Namely, that GenAI is a General Purpose (enabling) Technology (GPT), but also demonstrates the characteristics of an invention of methods of inventions (IMI). The authors conclude that these two traits will *likely* to sustained productivity growth, but that growth will take place over an extended period of time. (Regardless of any AGI declarations.) That productivity growth is not guaranteed. Lots of hard implementation work ahead. Extremely important cautionary note in the conclusion—massive growth in data centers & electricity generation is no guarantee of future success. We must not ignore the examples of boom-bust tech cycles throughout history. I haven’t changed my tune since the early days of Project Maven: tactical urgency, strategic patience. https://lnkd.in/ecwXg26H

Our recent research on the impact of generative AI–based tools on developer productivity found that a massive surge in productivity is possible. But there are two main mitigating factors: Task complexity and developer experience. We also found that developers using gen AI–based tools were more than twice as likely to report overall happiness, fulfillment, and a state of flow. This promises to help employers retain talent amid a persistent talent squeeze. To conduct this research, we set up what is now a permanent lab for ongoing analysis. Our lab has more than 40 McKinsey developers located across the U.S. and Asia who have different amounts of software-development experience. So stay tuned. In the meantime, here’s the full report from our first exercise: https://mck.co/46ttJaX #GenerativeAI #TechTalent #DeveloperTools #ArtificialIntelligence

A skeptical and honest look at the GenAI industry in 2025 Has anyone else noticed the GenAI industry’s ongoing tug-of-war between hype and reality? Launch ambitious tools with catchy names, often buggy, fragile, or unfinished. Amplify through breathless marketing. Cue the chorus: “This changes everything! Did you see what OpenAI just dropped?” "OMG the world will never be the same!" "Hurry up or get left behind!" Let’s pause. Let’s get real. I'm so over it to be honest The transformer architecture hasn’t changed dramatically since 2017. But it has evolved: mixture-of-experts (MoE), longer context windows, better fine-tuning. The result? Higher average performance, but also more hallucinations and less reliability in edge cases.¹ Agentic AI? Fascinating, sure. Today, 25% of GenAI-using companies are piloting agentic systems. That figure is projected to hit 50% by 2027.² Yet over 40% of these pilots are expected to be canceled due to cost overruns, fragility, and unclear ROI.³ CISOs are right to be cautious, given the mounting security and governance risks.⁴ In education, GenAI has potential, accessibility tools, engagement boosts, and personalized learning.⁵ But serious research from MIT, Stanford, and RAND warns of deep concerns: AI-driven misinformation, algorithmic bias, and potential cognitive decline linked to overreliance.⁶ Ethical governance in this space remains critically underdeveloped. In business, the Gartner Hype Cycle puts us in the “trough of disillusionment.”⁸ 78% of enterprises have run GenAI pilots, yet the same proportion reports no significant earnings impact.⁹ Roughly 30% of those pilots will be abandoned by end-2025 due to lack of measurable ROI.¹⁰ CEOs are reassessing: yes, there are real productivity bumps (up to 66%) in targeted functions like customer service and data analysis.¹¹ And the job market? No AI-led apocalypse. While over 77,000 tech jobs were cut in H1 2025 citing automation,¹² macroeconomic factors and post-pandemic hiring reversals are more significant drivers.¹³ Net employment impact remains minimal.¹⁴ So where are we? This isn’t a revolution. It’s a refinement phase, one with enormous long-term potential, and equally enormous near-term noise. Can we stop treating every product update like the singularity just arrived? Can we stop pretending this tech is either savior or snake oil? I'm not anti-GenAI. I'm pro-truth. Pro-balance. Pro-vision. And in 2025, that’s finally starting to feel possible. ******************************************************************************** The trick with technology is to avoid spreading darkness at the speed of light Sign up: Curiouser.AI is the force behind The Rogue Entrepreneur, a masterclass series for builders, misfits, and dreamers. Inspired by The Unreasonable Path, a belief that progress belongs to those with the imagination and courage to simply be themselves. To learn more, DM or email stephen@curiouser.ai (LINK IN COMMENTS)