Enhance or Eliminate? How AI Will Likely Change These Jobs
The following contribution comes from the Harvard Business Review Working Knowledge portal: Distilling Harvard Business School research for leaders who make a difference.
It is authored by Suraj Srinivasan, Philip J. Stomberg Professor of Business Administration and Director of the MBA elective program. He is a member of the Accounting and Management faculty and Director of the Digital Value Lab at Harvard’s Institute for Design, Data & Digitization. He co-directs the Harvard Business School (HBS) MBA program as Director of the elective program.
Research by Suraj Srinivasan reveals that employers are increasingly seeking AI-related skills in certain fields, while the demand for structured and repetitive tasks is declining. Discover which roles will be enhanced or automated in this interactive graphic.
Harvard Business School
Working Knowledge
Improve or Eliminate? How AI Is Likely to Change These Jobs
Will generative AI replace your job or improve it? What has been the impact on the job market so far?
Following the public launch of ChatGPT in November 2022, job postings for occupations involving many structured and repetitive tasks—likely replaceable by generative AI—decreased by 13%. Meanwhile, demand for jobs requiring greater analytical, technical, or creative skills—potentially enhanced by artificial intelligence—grew by 20%, according to a working paper co-authored by Suraj Srinivasan, a professor at Harvard Business School.
The findings offer early insights into how companies are adopting generative AI, which has led to a quest for efficiency within corporations and some unease among employees. The research team analyzed job postings from 2019 through March 2025 using a comprehensive dataset that covers nearly all job openings in the United States.
“Rather than simply eliminating jobs, generative AI creates new demand in positions that can be improved, suggesting that human-AI collaboration is a key driver of labor market transformation,” says Srinivasan. The largest reductions were seen in the financial and technology sectors.
Srinivasan, the Philip J. Stomberg Professor of Business Administration, collaborated on the working paper “Displacement or Complementarity? The Impact of Generative AI on the Labor Market” with Wilbur Xinyuan Chen of the Hong Kong University of Science and Technology and Saleh Zakerinia of The Ohio State University. The paper was first published in December 2024 and updated in August.
Professions with potential for AI integration encompass tasks that can be automated using generative AI, along with others that require human intervention. Those most susceptible to this integration tend to involve greater use of soft skills and practical techniques. Microbiologists, financial analysts, and clinical neuropsychologists are three examples with high potential for integration. In finance, as Srinivasan explains, investment managers and analysts use AI-based tools to process and evaluate market data, but ultimately, their judgment and decision-making ability remain crucial.
The research team used OpenAI’s ChatGPT to categorize more than 19,000 job tasks across more than 900 occupations, assessing their automation potential using generative AI. They also developed a augmentation score based on the proportion of exposed and unexposed tasks in each occupation.
Researchers found that the number of skills required for jobs susceptible to automation is decreasing. They recorded 7% fewer of these skills in job postings and also a lower number of emerging skills in these occupations. At the same time, they detected more AI-related skills—such as writing instructions or using AI tools—in jobs with high automation potential. As workflows are transformed by new technology, new skills have also emerged.
The researchers note that the study focuses on the short-term impact of generative AI on the US labor market, so the effects in other regions or the long-term impacts “remain uncertain as its adoption becomes more widespread.”
The report warns that how companies integrate generative AI technologies is crucial for job losses or growth. Given the varying impact on jobs, Srinivasan recommends that companies:
Invest in reskilling programs to facilitate the transition of workers to AI-enhanced roles. “Reskilling is essential for jobs where generative AI is reducing skill diversity. In occupations susceptible to automation, workers may face job displacement unless they develop non-automatable skills, such as judgment and interpersonal communication skills.” Provide ongoing training in generative AI to leverage new tools. “In occupations requiring enhancements, generative AI is expanding skill requirements, increasing the demand for AI expertise, human-AI collaboration, and industry-specific AI applications.”
“Companies should view generative AI as an enhancement tool, rather than simply a cost-cutting measure, and adapt staff training programs to support both job transitions and evolving skills demands,” Srinivasan says.
AI for the Entire Organization: Specialists, Generalists, and Executives
The following contribution comes from the Correlation One portal, which describes itself as follows: We are democratizing access to essential skills.
Digital and data literacy is fundamental to the future of work. However, structural barriers persist that hinder equitable access to essential training. This creates a growing skills gap within organizations and a sense of stagnation among workers.
In response, we are democratizing access to key skills through company-sponsored training, ensuring that everyone can thrive in the digital age.
Authorship by the team.
As artificial intelligence (AI) continues to redefine the boundaries of technology and business, it is becoming increasingly essential for organizations to develop AI readiness. But what does this readiness entail, and who is responsible for driving it?
The answer lies in understanding that AI readiness is not an individual task, but a collective effort, best achieved through the balanced integration of three key personnel types: specialists, generalists, and executives. Each plays a vital role in the adoption and implementation of AI. Specialists, with their deep technical knowledge, act as the essential cogs in the AI machine, managing tasks ranging from advanced programming to data analysis. Generalists, meanwhile, serve as indispensable bridges, connecting the specialists’ technical world with the executives’ strategic perspective. Finally, executives, with their strategic vision and leadership, provide the necessary direction and create an environment conducive to innovation.
Collaboration among these three roles forms the basis of AI readiness, fostering a culture of innovation and ensuring the effective use of AI technologies. In this article, we will delve into the roles, responsibilities, and skills that each of these profiles brings, as well as the need for harmonious collaboration among them and ways to cultivate this talent at all levels.
AI Specialists Lay the Foundation
An AI specialist is a person with extensive technical expertise who knows how to develop, use, and implement AI technologies and platforms. They may have designed AI algorithms and models or implemented enterprise-specific AI systems and tools. AI specialists include data scientists, machine learning engineers, software engineers, and AI researchers.
Specialists typically have backgrounds in mathematics, computer science, advanced programming, statistics, statistical modeling, or industry knowledge. The daily tasks of AI specialists may include developing and testing AI algorithms and models, collaborating with an organization’s multidisciplinary teams on AI implementation, and researching, implementing, and monitoring new AI technologies over time. Specialists also program AI systems, meaning they create specific instructions or tasks for their organization that the AI will execute.
AI specialists play a crucial role in the evolution of AI within the enterprise, but their role is not without its challenges. Due to their deep technical knowledge, specialists may find it difficult to explain complex AI concepts, algorithms, or models to team members or managers without technical expertise. They may also sometimes find it challenging to align their technical work with the overall business strategies defined by management.
To overcome these challenges, organizations should:
Emphasize data storytelling. Data storytelling is a powerful communication tool that combines data, visuals, and narrative to explain complex AI concepts or perspectives in a more accessible way. Encouraging AI specialists to use this approach when presenting their work to non-technical stakeholders can help bridge the understanding gap and make AI projects more relevant to overall business objectives. Consider conducting workshops or hiring external experts to train your teams in effective data storytelling techniques.
Foster cross-functional collaboration. Create opportunities for regular interactions and meetings among specialists, generalists, and executives. This could include project teams, brainstorming sessions, and strategic discussions. Such interactions can foster mutual understanding, promote knowledge sharing, and lead to more synergistic decision-making.
Establish clear communication channels and processes. Develop a robust communication framework that ensures a smooth flow of information at all levels. This could include regular status updates, project dashboards, and feedback mechanisms. Ensure that technical teams articulate their ideas and findings in a way that non-technical stakeholders can understand, and vice versa.
AI Generalists Bridge the Gap
AI generalists are team members who bridge the gap between technical AI roles and business-centric functions within the company. They may not have the specialized skills to develop or program AI, but they need to understand how to interact with AI-generated information and leverage it in their work.
Generalists often have backgrounds in non-technical fields, such as marketing or accounting, but they use AI to streamline their tasks. For example, a finance generalist might use AI-based predictive modeling tools to analyze large volumes of transactional data for unusual patterns or anomalies. In this role, the AI generalist wouldn’t necessarily develop these machine learning models, but would instead apply these tools, interpret the results, and translate these insights into practical risk mitigation strategies.
AI generalists often occupy a privileged position within a company, operating at the intersection of technology and business. While this provides them with a comprehensive view of the organization, it also presents several challenges. AI generalists may find it difficult to fully grasp the technical details of AI solutions, given their breadth of knowledge compared to the depth of AI specialists. Similarly, generalists may struggle to translate complex AI concepts or results into information easily understood by executives or other team members without technical expertise.
There are several steps organizations can take to overcome these challenges and fully leverage the potential of AI generalists within the organization:
Prioritize data literacy. Since AI and data are intrinsically linked, a deep understanding of data is vital for anyone involved in AI projects. Therefore, employers should prioritize data literacy among their generalists, which includes understanding how to read, work with, analyze, and interpret data. Employers can offer training sessions, workshops, or online courses to develop these skills. By promoting data literacy, organizations empower their generalists to understand the nuances of AI-generated information, contribute effectively to data-driven decisions, and collaborate more efficiently with AI specialists.
Establish clear communication channels. Ensure clear communication channels exist between AI specialists and generalists. This could include regular updates, feedback sessions, and an open-door policy for inquiries and clarifications. Good communication can prevent misunderstandings and ensure that all team members are on the same page.
To overcome these challenges, organizations should:
Emphasize data storytelling. Data storytelling is a powerful communication tool that combines data, visuals, and narrative to explain complex AI concepts or perspectives in a more accessible way. Encouraging AI specialists to use this approach when presenting their work to non-technical stakeholders can help bridge the understanding gap and make AI projects more relevant to overall business objectives. Consider conducting workshops or hiring external experts to train your teams in effective data storytelling techniques.
Foster cross-functional collaboration. Create opportunities for regular interactions and meetings among specialists, generalists, and executives. This could include project teams, brainstorming sessions, and strategic discussions. Such interactions can foster mutual understanding, promote knowledge sharing, and lead to more synergistic decision-making.
Establish clear communication channels and processes. Develop a robust communication framework that ensures a smooth flow of information at all levels. This could include regular status updates, project dashboards, and feedback mechanisms. Ensure that technical teams articulate their ideas and findings in a way that non-technical stakeholders can understand, and vice versa.
AI Generalists Bridge the Gap
AI generalists are team members who bridge the gap between technical AI roles and business-centric functions within the company. They may not have the specialized skills to develop or program AI, but they need to understand how to interact with AI-generated information and leverage it in their work.
Generalists often have backgrounds in non-technical fields, such as marketing or accounting, but they use AI to streamline their tasks. For example, a finance generalist might use AI-based predictive modeling tools to analyze large volumes of transactional data for unusual patterns or anomalies. In this role, the AI generalist wouldn’t necessarily develop these machine learning models, but would instead apply these tools, interpret the results, and translate these insights into practical risk mitigation strategies.
AI generalists often occupy a privileged position within a company, operating at the intersection of technology and business. While this provides them with a comprehensive view of the organization, it also presents several challenges. AI generalists may find it difficult to fully grasp the technical details of AI solutions, given their breadth of knowledge compared to the depth of AI specialists. Similarly, generalists may struggle to translate complex AI concepts or results into information easily understood by executives or other team members without technical expertise.
There are several steps organizations can take to overcome these challenges and fully leverage the potential of AI generalists within the organization:
Prioritize data literacy. Since AI and data are intrinsically linked, a deep understanding of data is vital for anyone involved in AI projects. Therefore, employers should prioritize data literacy among their generalists, which includes understanding how to read, work with, analyze, and interpret data. Employers can offer training sessions, workshops, or online courses to develop these skills. By promoting data literacy, organizations empower their generalists to understand the nuances of AI-generated information, contribute effectively to data-driven decisions, and collaborate more efficiently with AI specialists.
Establish clear communication channels. Ensure clear communication channels exist between AI specialists and generalists. This could include regular updates, feedback sessions, and an open-door policy for inquiries and clarifications. Good communication can prevent misunderstandings and ensure that all team members are on the same page.
Support continuous learning. Foster a culture of continuous learning within the organization. Encourage generalist professionals to stay current with the latest AI trends and technologies. This could be facilitated by providing them with access to learning resources, sponsoring their attendance at relevant conferences, or granting them time off for self-directed learning. Continuous learning can help
