In the latter half of 2025, concerns surrounding artificial intelligence and its potential to displace jobs intensified, leading to widespread jokes about AI’s true objective being “wages.” However, a recent study from METR, a nonprofit organization focused on researching AI’s risks and impacts, suggested that these fears may be exaggerated. The research, which examined AI’s effect on productivity, particularly among software developers, revealed surprising results.
METR, which stands for Model Evaluation & Threat Research, aimed to gauge the productivity changes brought on by AI assistance. The organization instructed 16 experienced software developers to complete 246 tasks using advanced AI tools. This approach differed from “vibe coding,” a method where less skilled coders utilize large language models to generate code without comprehending the underlying logic.
The findings were unexpected. According to Fortune, the researchers and participants alike were taken aback by the results, which indicated that developers using AI tools actually took 19% longer to complete tasks than those who worked without them. This revelation raises questions about the true efficiency and effectiveness of AI in enhancing productivity within this tech-savvy industry.
While these findings may provide some comfort to professionals in jobs perceived as vulnerable to automation, they also highlight broader concerns about the potential formation of an “AI bubble.” As 2025 progressed, enthusiasm for AI began to wane amid growing worries that investment in AI technologies might be outpacing their actual value, leading to a potential market collapse. Prominent investor Michael Burry, known for predicting the 2008 housing market crash, voiced his apprehensions about the burgeoning AI bubble.
Throughout the second half of 2025, major technology companies made substantial, sometimes frantic, investments in AI without clear strategies for profitability. This trend coincided with rising public discontent over AI data centers, which have been increasingly criticized for their significant resource consumption. Many communities began protesting the construction of these facilities, which have been linked to rising electricity costs and strain on local resources like water and energy.
As demand for AI-driven services surged, Americans witnessed soaring electric bills, largely attributed to the energy consumption of data centers. The Department of Energy issued warnings about inadequate grid capacity and an escalating risk of blackouts, suggesting that the infrastructure may struggle to keep pace with the rapid expansion of AI technologies.
Economist Anders Humlum, who studies the intersection of AI and the workplace, framed METR’s findings as a testament to the value of human expertise. He emphasized the importance of experience in the field, cautioning against a blind reliance on AI tools. “Many experts have a lot of experience… We should not just ignore that,” he told Fortune. Humlum’s insights serve as a reminder of the need for careful consideration regarding when and how to utilize AI in professional settings.
Moving forward, METR plans to continue assessing the real-world productivity impacts of AI, highlighting the need for ongoing research in this rapidly evolving field. On an individual level, staying informed about the environmental implications of AI, particularly concerning data centers, and advocating for regulatory measures can help mitigate some of the adverse effects associated with their expansion.
As the AI landscape continues to evolve, the balance between technological advancement and its broader societal impacts remains a critical focal point for policymakers, businesses, and the public alike.
See also
AI Founders Misjudge Uniqueness as Industry Advances; Clear Communication Key to Distinction
Maxon Launches Digital Twin Tool Amid Backlash from 3D Artists Over AI Focus
AI Weather Forecasting Tools Underperform in Predicting Extreme Events, Study Finds
Top 10 AI Automation Tools for 2026: Streamline Workflows and Boost Efficiency
Doximity Valuation Remains Undervalued at $46.49 Despite 4.99% Gains and AI Tools Launch
