{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,13]],"date-time":"2026-09-13T02:17:48Z","timestamp":1789265868552,"version":"build-2803163510"},"posted":{"date-parts":[[2023]]},"group-title":"SSRN","reference-count":68,"publisher":"Elsevier BV","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"<jats:p>We introduce and study the concept of a \u201cjagged technology frontier\u201d to describe the uneven impact of artificial intelligence (AI) capabilities, where AI assistance improves performance for some tasks but worsens it for others, even within the same knowledge workflow and with a seemingly similar level of difficulty. In collaboration with the global management consulting firm Boston Consulting Group, we have developed realistic management consulting tasks and examined the human performance implications of using AI to perform complex and knowledge-intensive work. The preregistered experiment\n&lt;br&gt;\ninvolved 758 knowledge workers. After establishing a performance baseline on similar tasks, subjects were randomly assigned to one of three conditions: no AI access, GPT-4 AI access, or GPT-4 AI access with a prompt engineering overview. For each one of a set of 18 realistic knowledge tasks within the frontier of AI capabilities ranging from creative to analytical tasks, subjects using AI outperformed those not using AI, completing 12.2% more tasks and completing them 25.1% more quickly on average while also delivering solutions of significantly improved quality. However, for a complex managerial task selected to be outside the frontier, subjects using AI were 19% less likely to produce correct solutions compared with those without AI, pointing to potential limitations of AI supporting knowledge workers. We discuss the positive and negative implications of AI-aided human performance in knowledge-intensive tasks.<\/jats:p>","DOI":"10.2139\/ssrn.4573321","type":"posted-content","created":{"date-parts":[[2023,9,18]],"date-time":"2023-09-18T08:54:45Z","timestamp":1695027285000},"source":"Crossref","is-referenced-by-count":588,"title":["Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of Artificial Intelligence on Knowledge Worker Productivity and Quality"],"prefix":"10.2139","author":[{"given":"Fabrizio","family":"Dell'Acqua","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Edward","family":"McFowland III","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ethan R.","family":"Mollick","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3461-003X","authenticated-orcid":true,"given":"Hila","family":"Lifshitz-Assaf","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Katherine","family":"Kellogg","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Saran","family":"Rajendran","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lisa","family":"Krayer","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fran\u00e7ois","family":"Candelon","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5535-8304","authenticated-orcid":true,"given":"Karim R.","family":"Lakhani","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"key":"ref1","first-page":"1043","article-title":"Skills, tasks and technologies: Implications for employment and earnings","volume":"4","author":"D Acemoglu","year":"2011","journal-title":"Labor Economics"},{"key":"ref2","author":"A Agrawal","year":"2018","journal-title":"Prediction Machines: The Simple Economics of Artificial Intelligence"},{"issue":"5","key":"ref3","doi-asserted-by":"crossref","first-page":"1090","DOI":"10.1227\/neu.0000000000002551","article-title":"Performance of ChatGPT, GPT-4, and Google Bard on a neurosurgery oral boards preparation question bank","volume":"93","author":"R Ali","year":"2023","journal-title":"Neurosurgery"},{"issue":"1","key":"ref4","doi-asserted-by":"crossref","first-page":"149","DOI":"10.1287\/orsc.2021.1554","article-title":"Algorithm-augmented work and domain experience: The countervailing forces of ability and aversion","volume":"33","author":"R Allen","year":"2022","journal-title":"Organ. Sci"},{"issue":"5","key":"ref5","doi-asserted-by":"crossref","first-page":"1672","DOI":"10.1287\/orsc.2022.1651","article-title":"Collaborating\" with AI: Taking a system view to explore the future of work","volume":"34","author":"C Anthony","year":"2023","journal-title":"Organ. Sci"},{"issue":"3 Part 2","key":"ref6","doi-asserted-by":"crossref","first-page":"849","DOI":"10.1287\/isre.1110.0408","article-title":"Information, technology, and information worker productivity","volume":"23","author":"S Aral","year":"2012","journal-title":"Inform. Systems Res"},{"issue":"1","key":"ref7","doi-asserted-by":"crossref","first-page":"80","DOI":"10.1257\/pandp.20201034","article-title":"The allocation of decision authority to human and artificial intelligence","volume":"110","author":"S Athey","year":"2020","journal-title":"AEA Papers Proc"},{"issue":"1","key":"ref8","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1287\/orsc.2021.1562","article-title":"We are all theorists of technology now: A relational perspective on emerging technology and organizing","volume":"33","author":"D E Bailey","year":"2022","journal-title":"Organ. Sci"},{"issue":"5","key":"ref9","doi-asserted-by":"crossref","first-page":"1448","DOI":"10.1287\/orsc.1100.0639","article-title":"Reconfiguring boundary relations: Robotic innovations in pharmacy work","volume":"23","author":"M Barrett","year":"2012","journal-title":"Organ. Sci"},{"issue":"1","key":"ref10","doi-asserted-by":"crossref","first-page":"87","DOI":"10.1177\/0001839217751692","article-title":"Shadow learning: Building robotic surgical skill when approved means fail","volume":"64","author":"M Beane","year":"2019","journal-title":"Admin. Sci. Quart"},{"issue":"5","key":"ref11","doi-asserted-by":"crossref","first-page":"2025","DOI":"10.1111\/joms.12867","article-title":"Pace layering as a metaphor for organizing in the age of intelligent technologies: Considering the future of work by theorizing the future of organizing","volume":"62","author":"M I Beane","year":"2025","journal-title":"J. Management Stud"},{"issue":"4","key":"ref12","doi-asserted-by":"crossref","first-page":"424","DOI":"10.5465\/amd.2023.0106","article-title":"Capturing value from artificial intelligence","volume":"9","author":"J M Berg","year":"2023","journal-title":"Acad. Management Discoveries"},{"key":"ref13","article-title":"The rapid adoption of generative AI","author":"A Blandin","year":"2026","journal-title":"Management Sci"},{"issue":"7992","key":"ref14","doi-asserted-by":"crossref","first-page":"570","DOI":"10.1038\/s41586-023-06792-0","article-title":"Autonomous chemical research with large language models","volume":"624","author":"D A Boiko","year":"2023","journal-title":"Nature"},{"issue":"5","key":"ref15","doi-asserted-by":"crossref","first-page":"1589","DOI":"10.1287\/orsc.2023.18430","article-title":"The crowdless future? Generative AI and creative problemsolving","volume":"35","author":"L Boussioux","year":"2024","journal-title":"Organ. Sci"},{"issue":"4","key":"ref16","doi-asserted-by":"crossref","first-page":"217","DOI":"10.1057\/s11369-021-00224-5","article-title":"The power of prediction: Predictive analytics, workplace complements, and business performance","volume":"56","author":"E Brynjolfsson","year":"2021","journal-title":"Bus. Econom"},{"issue":"2","key":"ref17","first-page":"889","article-title":"Generative AI at work. Quart","volume":"140","author":"E Brynjolfsson","year":"2025","journal-title":"J. Econom"},{"issue":"1","key":"ref18","doi-asserted-by":"crossref","first-page":"43","DOI":"10.1257\/pandp.20181019","article-title":"What can machines learn and what does it mean for occupations and the economy?","volume":"108","author":"E Brynjolfsson","year":"2018","journal-title":"AEA Papers Proc"},{"key":"ref19","author":"Z Buc \ufffdinca","year":"2021","journal-title":"To trust or to think: Cognitive forcing functions can reduce overreliance on AI in"},{"issue":"1","key":"ref20","article-title":"Navigating the Jagged Technological Frontier Organization Science, Articles in Advance","volume":"5","author":"Dell'acqua","journal-title":"Proc. ACM Human-Comput. Interaction"},{"issue":"2","key":"ref21","article-title":"Deep learning for chest X-ray analysis: A survey","volume":"73","author":"E Sogancioglu","year":"2021","journal-title":"Medical Image Anal"},{"issue":"2","key":"ref22","first-page":"384","article-title":"AI assistance in legal analysis: An empirical study","volume":"73","author":"J H Choi","year":"2024","journal-title":"J. Legal Ed"},{"issue":"2","key":"ref23","first-page":"536","article-title":"Human-AI ensembles: When can they work?","volume":"51","author":"V Choudhary","year":"2023","journal-title":"J. Management"},{"key":"ref24","first-page":"679","article-title":"Biased programmers? Or biased data? A field experiment in operationalizing AI ethics","author":"B Cowgill","year":"2020","journal-title":"Proc. 21st ACM Conf. Econom. Comput. (EC '20)"},{"issue":"7887","key":"ref25","doi-asserted-by":"crossref","first-page":"70","DOI":"10.1038\/s41586-021-04086-x","article-title":"Advancing mathematics by guiding human intuition with AI","volume":"600","author":"A Davies","year":"2021","journal-title":"Nature"},{"key":"ref26","author":"F Dell'acqua","year":"2022","journal-title":"Falling asleep at the wheel: Human\/AI collaboration in a field experiment on HR recruiters. Working paper, Laboratory for Innovation Science"},{"issue":"4","key":"ref27","doi-asserted-by":"crossref","first-page":"951","DOI":"10.1162\/rest_a_01328","article-title":"Super Mario meets AI: Experimental effects of automation and skills on team performance and coordination","volume":"107","author":"F Dell'acqua","year":"2025","journal-title":"Rev. Econom. Statist"},{"issue":"6797","key":"ref28","article-title":"Why providing humans with interpretable algorithms may, counterintuitively, lead to lower decision-making performance","author":"T Destefano","year":"2022"},{"key":"ref29","author":"P F Drucker","year":"1959","journal-title":"Landmarks of Tomorrow: A Report on the New Post Modern World"},{"issue":"6702","key":"ref30","doi-asserted-by":"crossref","first-page":"1306","DOI":"10.1126\/science.adj0998","article-title":"GPTs are GPTs: Labor market impact potential of LLMs","volume":"384","author":"T Eloundou","year":"2024","journal-title":"Science"},{"issue":"4","key":"ref31","doi-asserted-by":"crossref","first-page":"771","DOI":"10.1177\/14761270221130253","article-title":"Strategic organization in the digital age: Rethinking the concept of technology","volume":"20","author":"S Faraj","year":"2022","journal-title":"Strategic Organ"},{"key":"ref32","volume":"10","author":"E W Felten","year":"2023","journal-title":"Occupational heterogeneity in exposure to generative AI"},{"issue":"7","key":"ref33","doi-asserted-by":"crossref","first-page":"611","DOI":"10.1038\/s42256-022-00512-5","article-title":"Bringing artificial intelligence to business management","volume":"4","author":"S Feuerriegel","year":"2022","journal-title":"Nature Machine Intelligence"},{"issue":"1","key":"ref34","doi-asserted-by":"crossref","first-page":"161","DOI":"10.1086\/699936","article-title":"AI and the economy","volume":"19","author":"J Furman","year":"2019","journal-title":"Innovation Policy Econom"},{"issue":"11","key":"ref35","doi-asserted-by":"crossref","first-page":"2724","DOI":"10.1002\/smj.3512","article-title":"Training with AI: Evidence from chess computers","volume":"44","author":"F Gaessler","year":"2023","journal-title":"Strategic Management J"},{"issue":"2","key":"ref36","doi-asserted-by":"crossref","first-page":"233","DOI":"10.1177\/05694345231169654","article-title":"ChatGPT has aced the test of understanding in college economics: Now what?","volume":"68","author":"W Geerling","year":"2023","journal-title":"Amer. Economist"},{"issue":"4","key":"ref37","doi-asserted-by":"crossref","first-page":"619","DOI":"10.1002\/smj.3569","article-title":"Decision authority and the returns to algorithms","volume":"45","author":"E Glaeser","year":"2024","journal-title":"Strategic Management J"},{"issue":"6","key":"ref38","doi-asserted-by":"crossref","first-page":"1977","DOI":"10.1287\/orsc.2023.18441","article-title":"The short-term effects of generative artificial intelligence on employment: Evidence from an online labor market","volume":"35","author":"X Hui","year":"2024","journal-title":"Organ. Sci"},{"issue":"1","key":"ref39","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.2414972121","article-title":"The unequal adoption of ChatGPT exacerbates existing inequalities among workers","volume":"122","author":"A Humlum","year":"2025","journal-title":"Proc. Natl. Acad. Sci. USA"},{"key":"ref40","author":"M Iansiti","year":"2020","journal-title":"Competing in the Age of AI: Strategy and Leadership When Algorithms and Networks Run the World"},{"key":"ref41","article-title":"Commentary on jagged capabilities of large language models","author":"A Karpathy","year":"2024","journal-title":"Twitter"},{"key":"ref42","author":"A Karpathy","year":"2025","journal-title":"2025 LLM year in review"},{"issue":"2","key":"ref43","doi-asserted-by":"crossref","first-page":"571","DOI":"10.1287\/orsc.2021.1445","article-title":"Local adaptation without work intensification: Experimentalist governance of digital technology for mutually beneficial role reconfiguration in organizations","volume":"33","author":"K C Kellogg","year":"2022","journal-title":"Organ. Sci"},{"issue":"1","key":"ref44","doi-asserted-by":"crossref","first-page":"181","DOI":"10.1287\/orsc.2020.1374","article-title":"Moving violations: Pairing an illegitimate learning hierarchy with trainee status mobility for acquiring new skills when traditional expertise erodes","volume":"32","author":"K C Kellogg","year":"2021","journal-title":"Organ. Sci"},{"issue":"3","key":"ref45","doi-asserted-by":"crossref","first-page":"1501","DOI":"10.25300\/MISQ\/2021\/16564","article-title":"Is AI ground truth really true? The dangers of training and evaluating AI tools based on experts' know-what","volume":"45","author":"S Lebovitz","year":"2021","journal-title":"MIS Quart"},{"issue":"1","key":"ref46","doi-asserted-by":"crossref","first-page":"126","DOI":"10.1287\/orsc.2021.1549","article-title":"To engage or not to engage with AI for critical judgments: How professionals deal with opacity when using AI for medical diagnosis","volume":"33","author":"S Lebovitz","year":"2022","journal-title":"Organ. Sci"},{"issue":"13","key":"ref47","doi-asserted-by":"crossref","first-page":"1233","DOI":"10.1056\/NEJMsr2214184","article-title":"Benefits, limits, and risks of GPT-4 as an AI chatbot for medicine","volume":"388","author":"P Lee","year":"2023","journal-title":"New England J. Medicine"},{"key":"ref48","author":"L Meincke","year":"2024","journal-title":"Using large language models for idea generation in innovation. Working paper, Operations, Information and Decisions"},{"key":"ref49","first-page":"267","article-title":"An approach toward design and implementation of distributed framework for astronomical big data processing","volume":"431","author":"R Monisha","year":"2021"},{"issue":"7956","key":"ref50","doi-asserted-by":"crossref","first-page":"259","DOI":"10.1038\/s41586-023-05881-4","article-title":"Foundation models for generalist medical artificial intelligence","volume":"616","author":"M Moor","year":"2023","journal-title":"Nature"},{"issue":"6654","key":"ref51","doi-asserted-by":"crossref","first-page":"187","DOI":"10.1126\/science.adh2586","article-title":"Experimental evidence on the productivity effects of generative artificial intelligence","volume":"381","author":"S Noy","year":"2023","journal-title":"Science"},{"key":"ref52","volume":"15","author":"Openai","year":"2023","journal-title":"GPT-4 technical report"},{"key":"ref53","author":"N G Otis","year":"2024","journal-title":"Global evidence on gender gaps and generative AI"},{"key":"ref54","author":"N Otis","year":"2024","journal-title":"The uneven impact of generative AI on entrepreneurial performance"},{"key":"ref55","author":"S Peng","year":"2023","journal-title":"The impact of AI on developer productivity: Evidence from GitHub Copilot"},{"key":"ref56","article-title":"Interview with Lex Fridman","author":"S Pichai","year":"2025","journal-title":"Lex Fridman Podcast #471"},{"issue":"1","key":"ref57","doi-asserted-by":"crossref","first-page":"192","DOI":"10.5465\/amr.2018.0072","article-title":"Artificial intelligence and management: The automation-augmentation paradox","volume":"46","author":"S Raisch","year":"2021","journal-title":"Acad. Management Rev"},{"key":"ref58","author":"S Randazzo","year":"2025","journal-title":"GenAI as a power persuader: How professionals get persuasion bombed when they attempt to validate LLMs"},{"key":"ref59","author":"S Randazzo","year":"2025","journal-title":") Cyborgs, centaurs and self-automators: The three modes of human-GenAI knowledge work and their implications for skilling and the future of expertise"},{"key":"ref60","author":"S Reed","year":"2022","journal-title":"A generalist agent"},{"key":"ref61","author":"Dell'acqua","journal-title":"Navigating the Jagged Technological Frontier"},{"key":"ref62","first-page":"55565","article-title":"Are emergent abilities of large language models a mirage?","volume":"36","author":"R Schaeffer","year":"2023","journal-title":"Adv. Neural Inform. Processing Systems"},{"issue":"5","key":"ref63","doi-asserted-by":"crossref","first-page":"1248","DOI":"10.1287\/orsc.2019.1343","article-title":"Losing touch: An embodiment perspective on coordination in robotic surgery","volume":"31","author":"A V Sergeeva","year":"2020","journal-title":"Organ. Sci"},{"issue":"7972","key":"ref64","doi-asserted-by":"crossref","first-page":"172","DOI":"10.1038\/s41586-023-06291-2","article-title":"Large language models encode clinical knowledge","volume":"620","author":"K Singhal","year":"2023","journal-title":"Nature"},{"issue":"5","key":"ref65","doi-asserted-by":"crossref","first-page":"991","DOI":"10.1006\/ijhc.1999.0252","article-title":"Does automation bias decision-making?","volume":"51","author":"L J Skitka","year":"1999","journal-title":"Internat. J. Human-Comput. Stud"},{"issue":"12","key":"ref66","doi-asserted-by":"crossref","first-page":"2293","DOI":"10.1038\/s41562-024-02024-1","article-title":"When combinations of humans and AI are useful: A systematic review and metaanalysis","volume":"8","author":"M Vaccaro","year":"2024","journal-title":"Nature Human Behav"},{"key":"ref67","article-title":"MBB explained: How hard it is to get hired and what it's like to work for the prestigious strategy consulting firms","author":"K Vlamis","year":"2024","journal-title":"McKinsey, Bain, and BCG. Bus. Insider Africa"},{"issue":"3","key":"ref68","doi-asserted-by":"crossref","DOI":"10.1093\/pnasnexus\/pgae052","article-title":"Generative artificial intelligence, human creativity, and art","volume":"3","author":"E Zhou","year":"2024","journal-title":"PNAS Nexus"}],"container-title":[],"original-title":[],"deposited":{"date-parts":[[2026,8,14]],"date-time":"2026-08-14T17:42:58Z","timestamp":1786729378000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ssrn.com\/abstract=4573321"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023]]},"references-count":68,"URL":"https:\/\/doi.org\/10.2139\/ssrn.4573321","relation":{},"subject":[],"published":{"date-parts":[[2023]]},"subtype":"preprint"}}