Navigating Eight Paradoxes of AI for HR

AI, AI, AI, AI … all AI all the time.  In nearly every podcast, webinar, research report, and LinkedIn post (including this one), AI has become the topic du jour.  Using chatGPT, I learned that:

• AI-related startups are at record highs, with billions in funding. Some solve real problems—others chase hype.
• AI is now the most-mentioned term in S&P 500 earnings calls, more than “cloud” or “digital transformation” in previous years.
• Google Trends and LinkedIn posts around AI have surged 500–1000 percent in just two years (2023 to 2025).

Like others, I have tracked AI for HR movement up the s-curve of impact (see figure 1) with periodic posts:

• What AI offers as potential for HR.
• How to evolve AI for HR in four (now five) stages (access, assist, inform, guide, impact) (see figure 2).
• How to implement AI for HR using ten criteria.
• How AI can accelerate high-potential leaders.
• How AI delivers the next generation of stakeholder value creation through human capability.
• How AI helps prioritize where to focus to deliver stakeholder value.
• How AI can make more progress in upgrading HR.

Much has been done. Much more will come.

AI for HR has impacted divergent thinking, collaboration, and personalized actions, and has the potential to improve personal mental health, employee experience, leadership impact, organization results, and stakeholder value. For AI for HR to further deliver impact, our data suggest that paradoxes need to be navigated. Navigating (not just managing) paradox means highlighting and working through opposing ideas—each of which is valid—that combine to create more value. Navigating paradox includes: focusing less on “moving from . . . to” and more on “and . . .  also”; short and long term; individual competence and organization capability; local and global; empowerment and control.

Let me suggest eight paradoxes on the AI for HR agenda that business and HR leaders need to navigate to move up the s-curve (figure 1) and waves of HR impact (figure 2) to deliver more value. 

1. AI and AI: Artificial Intelligence * Authentic Intimacy

Perhaps the most prevalent AI for HR paradox combines technology (artificial intelligence) with humanity (authentic intimacy). Doing one without the other limits both (note the * in the formula, not a +). If at its foundation (figure 2), AI accesses information to make better decisions, some of that information comes from algorithms that rely on past data and some information comes from human perceptions of future opportunity. When algorithms combine with human empathy, judgement, and creativity, sustained progress occurs.

2. Remove Jobs and Redefine Work

Work is never a static noun but an evolving verb, constantly changing. AI will change how work is done, removing jobs that rely primarily on processing repetitive information (e.g., clerks, call center agents, cashiers, payroll, and other administrators) and redefining jobs requiring creative insight (e.g., teachers, doctors / medical workers, data analysts, trainers, and business and HR leaders).  Recognizing this paradox helps determine where and how to invest in AI vs. people.

3. Bottom Line Efficiency and Top Line Growth

Agentic AI, with bots doing work, can save time and money driving efficiency and reducing costs (e.g., processing resumes, personalizing training, compensation benchmarking, drafting and editing communications, etc.). GenAI can also provide information on how to grow revenue by increasing customer share, prioritizing human capability investments that deliver the most value, and anticipating future market opportunities. While AI helps reduce certain costs, its analytic insights also identifies which costs are the most impactful and worth continually investing in.

4. Distribute and Concentrate Power

AI empowers people by distributing information quickly and openly. Any internet user can use AI to access information. However, concentration of power from AI comes when those who use AI turn artificial intelligence into informed intelligence. These individuals access the right information, turn information into action, have courage to act, and exercise wisdom and judgment about what’s next.

5. Lower and Increase Risk

AI lowers risk by using algorithms to sort through volumes of data with structured analytics to determine patterns (reducing variance on the s-curve in figure 1). On the other hand, AI increases risk when it relies only on structured empirical analytics to make decisions. Combining structured data with unstructured data based on observation, experience, and intuition often leads to more effective decisions and higher analytical impact.

Risk also occurs by relying solely on AI data that may be wrong or fraudulent. Not all AI answers are right. Many have done AI searches on their work only to find that the responses are incomplete or wrong. I have had chatbots comment on my posts with misleading comments. The risk of AI misinformation goes up if not guided by human interaction. An estimated 9-15 percent of twitter accounts are bots, and they produce 20-25 percent of the total tweets.

6. Expand Perspective and Reduce Cognition

Those who regularly use AI can quickly access nearly unlimited information that increases perspective. For this post, I have learned about AI trends and percent of bot-driven tweets through chatGPT. I appreciate how easy getting answers to questions is. At the same time, we risk reduced cognition when relying solely on AI for information and insights. MIT research (Media Lab led by Nataliya Kosmyna) reveals that extensive use of AI reduced cognition function. The human contribution to problem solving enhances creativity, wisdom, and judgment in decision making that may be diminished by relying solely on AI generated information. 

7. Provide Answers and Explore Questions

AI quickly provides answers, or solves puzzles, which helps sort out what to think and do. AI also needs to pose questions and explore mysteries about what could happen. Answers often come from summarizing what has been done in the past; questions explore what could be done in the future. Using AI can do both when the answers received lead to further questions that lead to new actionable insights. 

8. Isolate and Connect

AI distributes work whenever and wherever a user desires: in an office, at home, at a coffee shop, hotel, or airplane. This personalized workspace may increase productivity and also a sense of loneliness and isolation. AI also has the ability to connect people around the world.  I write LinkedIn articles in my personal space, and the comments and engagement create connections globally.

Conclusion and Next Steps (another paradox):

AI, AI, AI, AI  . . . all AI all the time:

• Artificial Intelligence
• Authentic Intimacy
• Analytics Impact
• Actionable Insights

AI for HR is here to stay and will continue to evolve. Navigating the paradoxes of AI (in general and for HR) will help business and HR leaders be more conscious and intentional about how they use AI for impact (moving up the s-curve in figure 1 through the five waves in figure 2).

 The diagnostic in figure 3 allows assessment of current and desired state with four steps:

1. Step 1: Select the two to four most relevant AI for HR paradoxes for your company.
2. Step 2: Define the current focus and intensity (CBA to the left or 123 to the right).
3. Step 3: Identify where you would like your organization to be on this paradox (move left, right, or stay the same).
4. Step 4: Create forums (meetings, conferences, speeches, newsletters, podcasts, webinars, etc.) to identify and share desired changes and actions.

..………

Dave Ulrich is the Rensis Likert Professor at the Ross School of Business, University of Michigan, and a partner at The RBL Group, a consulting firm focused on helping organizations and leaders deliver value.