Leveraging Generative AI for Job Augmentation and Workforce Productivity 2024

Scenario 1 implications: – Organizations continue to develop GenAI, but the majority do so only within their internal organization for now. The risks are consider ed too high to integrate it into externally facingproducts and services.– Administrative jobs will begin to be augmented. Broader productivity gains will be minimal andmainly seen on an individual or team level. – Organizations will not systematically scale up GenAI for job augmentation, at least in the shortterm. Low trust, current applicability & quality The major difference between scenario 1 and 2 is the lack of enthusiasm and willingness among employees to deploy GenAI in scenario 2. Workers are inclined not to emphasize its potential and what it can do but rather the biases and the sometimes unreliable outcomes. This feeling is reinforced by stories about the unreliability of GenAI. Of course, there are employees in these organizations who do see the potential of GenAI and who do experiments, but they are too few to get their colleagues to go along on a large scale. Workers in this scenario may have access to the tools but tend to trust the quality of human work and judgement over technology so they do not feel motivated to try them out. Without humans-in-the-loop as well as policy there will be lower trust. This means that they do not experience the benefits GenAI can offer. GenAI is thus primarily used for labour-intensive and low-risk tasks like drafting an email. Employees are unlikely to trust the outcomes and instead conduct additional reviews of the outputs, often resulting in redoing the work. To increase adoption, organizations invest heavily in change management and training, but even these efforts do not fully overcome the resistance to embracing GenAI. In this scenario, the risks of “accidents” are not realized, but neither are the opportunities for innovation and job augmentation. Employees continue to manually handle most tasks that could have been automated or augmented by GenAI. Slight efficiency gains are achieved by performing simple tasks with GenAI, but the time required to verify the accuracy of outputs negates any potential benefits. External competitive pressures may slightly increase adoption, resulting in a marginally higher but still limited impact on productivity. In this scenario, adoption could be driven by external factors, such as competitive dynamics, despite its limitations in applicability and quality. Some companies might take the risk of adopting GenAI to gain an (uncertain) competitive edge. (Netflix, for instance, has expressed concerns regarding the potential impact of competitors leveraging GenAI technology on their ability to compete effectively). 41 Additionally, the desire to future-proof the organization, preparing for potential improvements in GenAI, might motivate organizations to start integrating the technology now. Scenario 2 implications: – Adoption of GenAI is very slow, thereby limiting the impact of GenAI on job augmentation and workfor ce productivity. – Individuals will use GenAI, but not on a large scale. Light individual productivity gains couldbe achieved by performing simple tasks withGenAI, but the effects will be negated by thetime required to verify the accuracy of outputs. – External pressur es may slightly increase adoption, resulting in a marginally higherbut still limited impact on augmentation andproductivity.Scenario 2: Broken Promises 2.2 13