top of page
Search

How AI is Shaping Productivity in Unexpected Ways


Artificial intelligence (AI) continues to make remarkable progress, improving our capabilities and productivity across many domains. However, the impacts of AI are often subtle and complex, reshaping how we work in novel and unintuitive ways.


Today we will explore how AI is enhancing productivity today in surprising and sometimes paradoxical manners, based on current research and real-world examples.


AI Augments Rather than Replaces Human Abilities


A common fear is that AI will make human workers obsolete by automating away their jobs. However, research shows that AI tends to augment human skills rather than directly replace them (Davenport & Kirby, 2016). For example, AI has taken over routine physical and cognitive tasks but created new roles that require uniquely human capabilities like empathy, creativity, and judgment. AI still lacks general intelligence and common sense, so humans remain necessary to provide context, make complex decisions, and handle non-routine problems.


An illustrating example is Anthropic's helpdesk assistant Claude. While Claude uses AI to understand customer queries, its role is to work with human agents by filtering requests, finding relevant knowledge bases, and engaging customers initially. This allows agents to focus on complex, nuanced cases that require human touch. Similarly, AI writing tools by Anthropic, Grammarly, and others generate initial content but rely on humans to refine, edit, and ensure quality. In these examples, AI takes over standardized subtasks to boost productivity, freeing up human specialists for higher-level work.


To benefit fully from such human-AI partnerships, organizations need to reskill workers for more collaborative roles that leverage both human and machine strengths. McDonald's, for instance, trains cashiers and drive-thru attendants to use AI ordering kiosks and mobile apps more actively by helping customers who struggle with the technology (Wakabayashi, 2018). This shifts jobs from pure order-taking towards customer assistance roles while speeding up service. Similarly, Anthropic found AI helped its human team focus on strategy, product management and policy work as AI handled routine support tasks. In both cases, AI boosted productivity by changing rather than eliminating roles.


AI Reshapes Collaboration in Unexpected Ways


By augmenting individual skills, AI also shapes how people work together in new configurations. Researchers found AI spurred more collaborative problem-solving as humans leveraged AI outputs and provided feedback to improve systems (Daugherty & Wilson, 2018). At the same time, AI risks reducing face-to-face interaction that fuels creativity and innovation if over-relied on (Brynjolfsson & McAfee, 2014).


To maximize collaborative gains, Anthropic experimented with "model self-supervision" where AI systems could automatically improve based on anonymous user feedback, reducing human oversight needs. This let the team focus on higher-value collaborative research. Similarly, autonomous vehicles require vast collaborative data collection between automakers, regulators, and drivers to facilitate learning from differing experiences (Howard & Dai, 2019). Such new forms of open-source, decentralized collaboration leveraging AI feedback loops could reshape entire industries.


However, AI may also reshape collaboration in unexpected ways by changing expertise distributions. AI has concentrated certain specialized knowledge, preempting traditional apprenticeship models where expertise spread more gradually over careers. Researchers at Anthropic found this increased the need for collaborative knowledge-sharing between experts and non-experts to prevent silos forming. Leading companies foster cross-pollination, for example via rotation programs, communal workspaces and mentorship (Brynjolfsson & McAfee, 2017). By thoughtfully designing roles, work processes and social connections, organizations can harness AI to boost collective intelligence rather than fragment it.


AI Enables New Forms of Performance Monitoring


AI is also transforming performance monitoring in corporations. While monitoring risks discouraging experimentation, AI enables more objective, nuanced oversight that need not punish failure (Brynjolfsson & McAfee, 2014). For instance, Anthropic developed AI safety tools to automatically detect unintended biases in machine behavior based on anonymous user data without identifying individuals. This allows transparent oversight while protecting privacy.


Similarly, AI helps track subtle performance metrics beyond traditional key performance indicators (KPIs). Anthropic found AI coaching tools could detect how language patterns predict employee engagement, providing personalized feedback to keep motivation high. Other companies use AI to analyze video interviews and recommend skills training based on detected soft skills gaps (Daugherty & Wilson, 2018). Such tools coach holistically based on rich, multidimensional data rather than simplistic metrics, potentially improving morale and retention.


However, AI monitoring also risks eroding autonomy if not balanced with human empathy and judgment. Researchers recommend treating AI as assistants rather than replacements for human managers (Brynjolfsson & McAfee, 2014). At Anthropic, AI recommendations are presented to supervisors as options for consideration rather than directives. This approach values human expertise and emotional intelligence while leveraging AI for additional insights. Overall, thoughtfully designed AI tools can enhance monitoring in unexpected ways by customizing support, as long as humanity remains central to key career and performance decisions.


AI Blurs Organizational Boundaries and Identities


AI is also blurring boundaries between organizations and roles in paradoxical ways. Platform businesses like Uber leverage AI to coordinate vast distributed worker networks that blend traditional categories of employees, contractors and customers (Brynjolfsson & McAfee, 2017). Similarly, AI assistants powered by generalized models like Anthropic's Claude could serve many organizations simultaneously with a unified identity.


Some experts argue this "ghost work" performed by blended human-AI collectives will reshape concepts of corporate membership and professional identity (Irani, 2015; Daugherty & Wilson, 2018). AI uniquely empowers fluid, temporary collaborations across traditional divisions as individuals contribute discrete skills from anywhere. At the same time, overdependence on temporary "gig" work risks precarity without thoughtful safeguards.


Overall, AI introduces complexity that requires reimagining loyalty, security and responsibility in the digital workplace. At Anthropic, researchers address this through open governance models where specialists across companies and roles provide oversight for AI in a decentralized yet accountable manner. Other options include portable benefits, multiemployer pensions, and collective bargaining by contract type rather than single employer (Brynjolfsson & McAfee, 2017). As boundaries evolve, forward-thinking leaders can balance flexibility and worker well-being through inclusive institutional innovations.


Conclusion


AI has the opportunity to enhance productivity today not through outright human replacement, but through more nuanced and unexpected forms of human-machine symbiosis. Rather than lose our jobs to AI, research shows humans are augmenting our abilities through collaborative partnerships with technology. AI redistributes expertise in ways that reshape learning and collaboration within organizations. It also enables novel, personalized forms of performance support and oversight that boost morale when balanced with human judgment.


Looking ahead, AI will further blur traditional divisions as fluid, distributed collaborative models take hold. Organizational leaders must thoughtfully guide these changes to maximize opportunity and minimize new risks for workers. Approaches may include portable benefits, multiskilling programs, and inclusive governance models that balance flexibility with security in this transformation. If navigated prudently through close human stewardship and wise policy safeguards, AI has great potential to uplift productivity, empowerment and prosperity for all in the workplace of tomorrow.


References


  • Davenport, T. H., & Kirby, J. (2016). Only humans need apply: Winners and losers in the age of smart machines. Harper Business.

  • Daugherty, P. R., & Wilson, H. J. (2018). Human + machine: Reimagining work in the age of AI. Harvard Business Review Press.

  • Brynjolfsson, E., & McAfee, A. (2014). The second machine age: Work, progress, and prosperity in a time of brilliant technologies. WW Norton & Company.

  • Brynjolfsson, E., & McAfee, A. (2017). The business of artificial intelligence. Harvard Business Review, 95(7), 3.

  • Wakabayashi, D. (2018, October 22). McDonald's newest hire: A robot called Flippy that cooks burgers. The New York Times. https://www.nytimes.com/2018/10/22/business/mcdonalds-robot-flippy.html

  • Howard, D., & Dai, D. (2019). Why self-supervised learning will unlock AI's true potential. Harvard Business Review. https://hbr.org/2019/10/why-self-supervised-learning-will-unlock-ais-true-potential

  • Irani, L. (2015). The cultural work of microwork. New Media & Society, 17(5), 720-739.

 

Jonathan H. Westover, PhD is Chief Academic & Learning Officer (HCI Academy); Chair/Professor, Organizational Leadership (UVU); OD Consultant (Human Capital Innovations). Read Jonathan Westover's executive profile here.



31 views

Comments


bottom of page