Steve starts season 3 of Integrate or Die interviewing co-host Alexis Fink about her experiences using AI technology as an HR leader in large, global organizations. She shares insights about the benefits and risks of using AI to change talent processes and decisions drawn from experience leading workforce analytics at Meta, Intel and Microsoft. Some things we cover:
+ Building “white box” vs “black box” solutions in a world where it is becoming easier to be misled by technology
+ The value of rethinking business processes to accommodate new technological capabilities
+ Why advancing your career often comes down to just doing new stuff on your own
We also learn that Alexis is a past president.
Steve and Alexis talk with Alec Levenson, Director and Senior Research Scientist at the USC Marshall Center for Effective Organizations about AI, data, and decision-making through the lens of organizational effectiveness. Alec brings an economist’s skepticism to the discussion, emphasizing the limits of data and technology when disconnected from real organizational questions. Some things we cover:
+ Why organizations repeatedly overestimate what data, analytics and AI alone can explain
+ If fears over AI and job loss are more about personal perspective than huge disruptions in the labor market
+ The danger of focusing on what technology can measure instead of what leaders need to know
We also learn from Alexis why we do not use AI to land planes even though we could.
Steve and Alexis talk with Armen Berjikly, an early-stage investor in technology solutions at Emerson Collective about integrating human involvement into AI automated processes. Armen is a computer scientist who built some of the first widely adopted natural language AI solutions applied to HR at Kenjoya (employee listening) and Better Up (coaching). Some things we cover:
+ The importance of investing in problem definition before investing in AI solutions
+ How AI can be used to understand and engage employees at all levels of the organization
+ Why we should resist the urge to give AI chatbots overly human personas
We also hear Steve explain why using the term “artificial intelligence” to describe LLMs is like using the term “energy drink” to describe caffeinated sugar water.
Steve and Alexis talk with Eric Sydell, founder & CEO of Vero AI about responsible use of AI solutions. Eric built some of the first large scale applications using AI machine learning to support recruiting as a founding member of Modern Hire, a company subsequently acquired by HireVue. He is now building solutions to ensure compliant use of AI applications across a range of areas. Some things we cover:
+ The mismatch between what leaders say about responsible AI and what they often do
+ The tendency for many companies to monitor AI bias without monitoring AI accuracy
+ How venture capital investment strategies negatively influence how AI solutions are built
We also learn how Alexis got a PhD from Harvard without ever attending a single class.
Evan Sinar, Ph.D., Senior Research Scientist at Amazon discusses how development of generative AI solutions is affecting psychological and organizational research. Evan led applied research teams at DDI and BetterUp before joining Amazon and works on the forefront of research focused on employee performance and development. He shares how AI is reshaping research methods but also calls out its limitations and risks. Some things we cover:
+ How the ability to analyze qualitative data is transforming organizational research
+ The largely untapped research knowledge available from mining legacy text sources
+ Why we should talk about time utilization instead of time savings when discussing the value of AI
We also hear Steve describe his career as a series of ostentatious sounding but vaguely meaningless job titles.
Steve and Alexis talk with Fred Oswald, Professor at the University of California Irvine School of Education about the impact of AI on the assessment and education of students and employees. This episode dives deep into some nerdy industrial-organizational concepts so be forewarned! Some things we cover:
+ Whether psychometric principles apply to the evaluation of AI/ML algorithms
+ How use of AI is changing the focus of higher education
+ The role of social sciences in an increasingly digitalized world
This is also the first episode that features not just one, but two past presidents of the Society for Industrial Organizational Psychology.
Steve and Alexis talk with Guy Tourigny, HR Technology Implementation Leader at HRIZONS about using AI to build agentic HR applications. We discuss the efficiency gains associated with these solutions along with the importance of not over automating a function that is focused on optimizing the value of humans. Some things we cover:
+ The value of agentic HR solutions that integrate data from different functional areas
+ Knowing when to switch from AI agents to human agents
+ Managing the risks when you rely on AI to interpret HR data policies instead of people
We also learn that Steve’s concepts of AI are rooted in how Lieutenant Uhuru used computers in the original Star Trek.
Steve and Alexis talk with Jerome Gouvernel, co-founder of datascalehr about the value AI provides for integrating highly complex, global HR datasets. Jerome’s work is primarily focused on payroll data management, which as we learn is one of the most important, most sensitive, and most complicated areas of HR data. Some things we cover:
+ Why HR data is fragmented and far more complex than people realize
+ How AI enables companies to align and harmonize imperfect data
+ The critical difference between predictive AI vs generative AI
We also hear Alexis describe data integration as the place where the “magic happens”.
Steve and Alexis talk with Richard Landers, Professor of Industrial Organizational Psychology at the University of Minnesota about building AI technology that is informed by psychological science. We focus specifically on the use of AI to support candidate selection and employee development. Some things we cover:
+ The change management challenges caused by the rapid speed of AI innovation
+ The challenge of building effective AI agents for tasks where effective human performance is poorly defined
+ The risk of doing things because they are new and different but not necessarily better
We also hear Steve share a tale from his childhood about his dad acting like a sensitive robot.
Steve and Alexis talk with Tim Gregory, VP of HR Transformation at Delta Air Lines about adoption and use of AI technology in large, complex global enterprise companies. We explore the challenges enterprise organizations face when using AI technology and discuss why the development of enterprise AI solutions follows a different timeline from consumer AI applications. Some things we cover:
+ The four levels of AI maturity within an enterprise organization
+ Why fully agentic AI solutions may be developed faster in enterprise companies than outside of them
+ Why skills management is a good training ground for using AI within an enterprise organization.
Alexis also tells us what it is like to confused with a generative AI solution whose name sounds like yours.
Steve and Alexis talk with Tyler Weeks, VP of People Technology, Research & Analytics at Marriott International about using AI to support frontline employees in the hospitality industry. A physicist by training, Tyler brings a high sophisticated but reality-based approach toward the use of AI technology within organizations. Some things we cover:
+ Why AI LLMs that use natural language interfaces should not be designed to give the illusion of actually being human
+ How use of AI is limited and often derailed by the complex and imperfect nature of organizational policies and processes
+ Why the next big wave of AI innovation may focus on using AI for understanding instead of automation
We also hear Tyler explain why he thinks of Alexis as a doula for workforce analytics professionals.