AI and HR: Rethinking How We Work with Diane Gherson, Member, Board of Directors at Kraft Heinz, Centivo, and TechWolf

The Next Practices Weekly call series has become a well-attended and wide-ranging discussion for HR leaders each Thursday at 11am ET / 8am PT. On this week's call, i4cp's Chief Research Officer, Kevin Martin, and Senior Research Analyst, Tom Stone, facilitated a conversation with special guest Diane Gherson, former CHRO at IBM, and now Member, Board of Directors at Kraft Heinz, Centivo, and TechWolf. Here are some highlights from the call:

Gherson began her career in HR with 11 years at Towers Watson, before joining IBM in 2002 as VP of Compensation and Benefits. She became the CHRO in 2013 and held that position until retiring from IBM at the end of 2020. She has been a Senior Lecturer at Harvard Business School, and serves on several corporate boards at organizations such as Kraft Heinz, Centivo, and TechWolf.
Gherson shared that several factors drove her interest in working more closely with IBM's board, and then to serve later on other organization's corporate boards as noted above. One was the #MeToo movement, and the need to change IBM's corporate code of conduct in a way that improved the culture, guarded the organization against brand damage, etc. COVID then came along and really led Boards of Directors to lean even more into issues around human capital, organization culture, etc.
Martin shared some recent i4cp research, done in conjunction with the group Extraordinary Women on Boards (EWOB), that found both a top-10 list of human capital data shared most often shared with full Boards of Directors (Diversity of Leadership and Leadership Bench / Succession were top of the list), and a top-10 list of human capital data NOT shared but that Board Directors would view as valuable (Culture Health and Internal Movement Rate were top of that list). For the full top-10 lists, see the slides or recording above.
In another recent study by i4cp and EWOB, Martin shared data that found only 9% of Board directors are very confident that their companies are effectively upskilling/reskilling employees for the future--versus 32% that were somewhat confident and 42% that were not confident. Similarly, only 16% of Board directors said they were very confident that their companies were communicating a vision about how generative AI might affect their organization or industry--versus 33% that were somewhat confident and 41% that were not confident.
Regarding AI, Gherson stressed the importance of involving employees, e.g., a council of managers, in the development of AI governance policies, because that gives them confidence that policies created aren't some sort of Orwellian control mechanism.
In using AI for sentiment analysis, Gherson said she looks for the themes, both the high notes and the low notes, in the employee feedback. This helps you zone in on issues, and strengths, in a particular region or large functional area--without violating privacy and anonymity in such employee surveys.
Gherson said that at IBM, their managers were overburdened, so they looked to AI to help alleviate some overload challenges. One example she gave was using AI to pull information from multiple systems in response to a manager's query, such as "What steps do I need to follow in order to terminate, transfer, or promote an employee?"
Another AI usage example was to give recommendations to managers with data on turnover rates, market demand data, etc., for particular skills, etc., to inform -- but not determine -- compensation / raise decisions. Over the years, the recommendations have gotten better, such that more and more managers can reasonably follow the advice given.
Gherson agreed with a participant's comment that one potential downside of generative AI is overuse of it in drafting job descriptions, training content, etc., such that it is obvious the content was AI-generated (which might not be well-received) and/or many organizations having very similar, and very generic content of this sort. AI, while ever improving, can also make mistakes--including hallucinating data/information that is entirely false. The solution is to see Generative AI as an assistant, as a tool to help draft, but not fully produce, finished content products.
Gherson also shared a worry she has about over-use of AI: if it replaces too many lower-level tasks, then is it eliminating the work that junior employees do as a stepping stone to develop their skills and later their careers in the organization?
When asked about forecasting future skills needed in an organization, Gherson said that any work on this must be aligned with business strategy. At IBM, they did this work once a year, identifying the skills they anticipate needing more of and less of going forward, and also connected these needs to the broader talent market. This included identifying skill adjacencies for employees to help them better chart their future.
Gherson also noted that humans remain stronger than AI when it comes to EQ - Emotional Intelligence. Generative AI will often not accurately ascertain subtleties in certain data sets where EQ is needed to pick-up on human nuances. So when hiring, selecting for strong EQ is as important as ever--and so is developing stronger EQ in current employees as well.

Note that Diane Gherson will also be joining i4cp at Next Practices Now as a featured speaker. Visit our conference website for all the 2024 speaker and agenda details.

Links to resources shared on the call:

About Next Practices Weekly

This series provides a forum for HR leaders to come together, discover, and advance cutting-edge human capital practices. Each week, you’ll hear top executives from some of the world's largest and most respected organizations explore the future of work.

This event is exclusively for HR practitioners. Vendors and consultants are not permitted to attend.