Let’s talk about how we talk about AI

Training and readiness programs aren’t exactly meeting the learning and development needs of employees or the business when it comes to AI. That’s the thesis, at least, for Howard Pyle, founder of XF, a think tank and tech developer focused on “responsible AI, both at the software layer and at the people layer.”

To help HR teams begin addressing this, Pyle’s XF released a new tool this week at HR Brew’s Talent 2030 event in New York City.

The new resource aims to help people take the reins of their AI use. It prompts users to explain their values and current AI-use using natural, non-technical language, translates those bespoke experiences into larger AI concepts to educate them on core AI abilities, and provides a “mini manifesto” they can equip their LLM with to help “learn us.”

“Most organizations are actually asking the wrong questions. I don’t think the question is: Have we deployed the right technology? Have we deployed the right technology training? And yes, you need both of those things. But what we really need to talk about is: How do we get our people to know themselves well enough to be able to use AI and to tell it exactly what they want it to do?” Pyle told more than 200 HR pros during the event. “We have a technology that’s supposed to learn us. We have to be able to tell it what we want it to do.”

This is different from the learning dynamics required for other technological transformations, Pyle said. This new technology requires HR and learning and development (L&D) pros to think differently about readying the workforce since it works better when it can “learn us,” rather than requiring users to learn how to engage with it to maximize its capabilities.

“The secret hint here is it’s not about technical ability. This is not a moment where dumping more technology on people is going to suddenly make them be more effective or more empowered in this era,” he said.

AI adoption requires developing human talent, not “to make them technologists,” but to express themselves and articulate their skills and what they want to get from AI or have it do for them.

“The most important thing of this AI era is being able to articulate your tacit knowledge to AI systems,” Pyle said. “But the truth is, when you put most people in front of a piece of paper and say, ‘Write down your unique expertise, your unique perspective, your unique way of seeing the world,’ they have a really, really hard time doing it.”

XF developed a “personal AI toolkit” as part of its broader Intentional Technology program that helps professionals create a personal strategy for AI in their career that can follow them as they use their different tools or assume different roles or jobs.

The resource asks a series of simple prompts about the human user’s work and where AI fits in. It then helps users outline and understand “AI constants” (a set of features and capabilities of AI, like “assisting,” “augmenting,” “deciding,” and even “representing”) and generates a “mini manifesto” that can be used to instruct LLMs like Gemini, ChatGPT, or internal agentic systems on the uses and guardrails.

“The idea is that if you can give people language for how to participate in this conversation, and you can coach them on understanding themselves, then they will be able to set guardrails, they will be able to say how they want AI to be work,” he said.