AI Activations - AI Literacy Beyond the Prompt: From Executor to AI Director

In this episode of AI Activations, Cobus Greyling sits down with Valentina Adami, a conversational AI and generative AI expert with over 20 years in IT, to unpack what AI literacy really means in 2026. Valentina argues that true literacy goes beyond knowing how to use an LLM; it requires understanding the probabilistic mechanics behind these systems, so people know not just what AI can do, but what it can't.

The conversation explores the "lights and shadows" of generative AI, from its transformative potential to its real-world failures, and dives into Anthropic's AI Fluency framework (Discovery, Description, Discernment, Diligence) as a practical lens for responsible adoption. Valentina distinguishes between a strategic approach (using AI to brainstorm and explore new possibilities) and a tactical approach (using AI to automate known, well-defined tasks), and shares candid insights from her work helping small and medium Italian enterprises adopt AI, including why many organizations default to bolting AI onto broken processes rather than redesigning them from the ground up.

Drawing on the "horseless carriage" analogy and lessons from IBM's historic process re-engineering, the discussion lands on a central theme: the shift from being an executor trained to follow instructions to becoming a director who orchestrates AI toward outcomes. It's a mindset shift Valentina believes is essential, and one that won't happen overnight.

Key topics covered:

• What AI literacy actually means today
• Anthropic's AI Fluency framework (the four Ds)
• Strategic vs. tactical use of AI
• Why SMEs struggle with AI adoption
• The "horseless carriage" trap of automating old processes
• Moving from execution to AI direction/orchestration

‍