人工智能如何看待自身反噬及对世界的影响

What AI says about AI eating itself andthe world... Jim Reid AI says it is targeting IT and software, finance, customer services,manufacturing and logistics, and media and entertainment Global Head of Macro and Thematic Research+44-20-754-72943 The market gyrations of the past two weeks, as fears about AI disruption hit firstsoftware and then technology and financial stocks, have left anxious investorswondering:which sector will be next to drop? Adrian CoxResearch Analyst+44-20-754-17775 We went straight to the source, asking ourproprietary AI tool dbLumina, whichruns on Google's Gemini 2.5 Pro, to do a Deep Research analysis. "You are a sophisticated analyst specialising in the implications of AI for theeconomy and markets," we said, asking it for "a report of no more than 3,000 wordswith deep analysis of which global sectors are likely to be most and least disruptedby AI". We specified a number of other parameters and, after some discussion, itgenerated the report you see below not much more than a minute later. AI is set to increase global GDP and labour productivity, the AI-generated reportsays.Data-rich sectors with repetitive, pattern-based tasksare most likely to bedisrupted, it says. Citing numerous sources from Stanford University, Brookings and the WorldEconomic Forum to Wikipedia, Reddit, blogger Brian Manning and one of ourcompetitors,it singles out information technology and software for disruption,thanks to automation of core tasks and disintermediation. (In the interest of transparency, we are publishing the report completely unedited.For readability, we have not included the links to sources in the original report.Please contact us if you have questions on sourcing.) Other industries ripe for disruption includefinance, thanks to the rise of robo-advisers and automation of back-office roles like data processing;customerservices;manufacturing and logistics; andmedia and entertainment, it says. Themost insulated sectorsrequireempathy and human connection, like directpatient care and education; those requiringmanual dexterity in unpredictableenvironmentssuch as construction and skilled trades; and those demandingstrategic and creative leadership, the report says. Corner-office executives andone-of-a-kind artists are safe for now. 16 February 2026Thematic Research We broadly agree with the conclusions,which are a faithful reflection of thecurrent consensus. Given how the model is trained—and the way it generates theoutput—it is unlikely to deliver genuinely novel insights. Still, it provides a usefulsynthesis of prevailing views, produced in seconds. That said,we would place more weight on the near-term obstacles to adoptionthan the report suggests.These range from data quality and governance tooperational constraints such as access to energy. Conversely, we think the long-term opportunity is likely even larger than implied—extending well beyond processautomation and improved analytics as genuinely new economic and businessmodels emerge. We also think the report goes too far in equating disruption with inherentlynegative outcomes.New technologies often enable enterprises and individuals towork faster, better, and more enjoyably. Many of the sectors and companies thathave sold off in recent days are likely to adapt and ultimately use these capabilitiesto their advantage, potentially becoming more productive and profitable. Markets,however, tend to sell first and reassess later. The report also cites headline figures—particularly around productivity gainsand employment impacts—that read as authoritative but may be drawn from asmall number of widely referenced studies.Those estimates can come todominate the literature through repetition rather than superior rigour, and the reportmay therefore unintentionally elevate their status. Beyond that, the report generated by dbLumina is well-structured and clear, evenif it inevitably lacks a memorable central argument and the writing is rather woolly.Not all of the information is supported by the links cited, and even where it is, thereare instances where two sources may be incompatible. Butit's a strong startingpoint, proving its own argument: AI is useful, potentially revolutionary, and hereto stay. See what you think. The Great Rebalancing: AnAnalysis of AI's DisruptiveImpact on Global Sectors An in-depth look at which industries are poised for AI-driven transformation andwhich are likely to remain bastions of human-centric work, with a focus onrecent market shifts in software and wealth management. Report generated by dbLumina in Deep Research mode, using Google's Gemini 2.5Pro AI model. Executive Summary Artificial Intelligence is a powerful disruptive force actively reshaping the globaleconomy,creating a great rebalancing where some sectors face profoundtransformation while others remain relatively insulated. nMost Likely to be Disrupted: Data-rich sectors with automatable tasks areat the forefront. These include Information Te