- Artificial Intelligence
- blog
The Scriptorium Initiative in collaboration with Page Society
Most conversations about AI focus on the future, asking what the technology might do next. Marc Zao-Sanders’ research asks a more relevant question: how are people actually using AI now?
Through his research initiative AI in the Wild, and in his recent Harvard Business Review article “How People Are Really Using AI in 2026,” Marc analyzed more than 12,000 real-world AI use cases. The result is a bottom-up view of AI adoption, focused less on hype and more on human behaviour.
What stands out is that AI is being used for much more than productivity. People use it to write, summarize, code, brainstorm, and work faster. Many of the most common uses, however, are far more personal. In 2026, the top category in Marc’s research was therapy and companionship. People are turning to AI as a sounding board, adviser, mentor, and source of emotional support.
That finding matters to businesses and CCOs. It suggests that for many people, AI is becoming a source of reassurance, reflection, entertainment, and guidance. Other popular uses, including storytelling, fanfiction, astrology, tarot readings, and imagined reality TV scenarios, show that adoption is also driven by play, curiosity, and emotional engagement.
For communications leaders, this insight is especially relevant. Generative AI is built around language, meaning, and interaction. It can draft and edit text, generate campaign ideas, summarize meetings, adapt messages for different audiences, and act as a strategic sounding board. The same capabilities used for productivity and speed also create risk.
One of Marc’s most useful concepts is “think slop.” AI slop refers to generic, low-value AI-generated content. Think slop describes what happens when people outsource the thinking process too early. Instead of forming their own view, they ask AI for an answer and accept something that is plausible, polished, and average. Think slop is defined by low quality of thought.
For communications, this is a serious warning. As we know and have experienced ourselves, writing is part of thinking, not only the final output. If AI takes over too much of the drafting process, it may weaken judgment, originality, and human ownership.
The question for organizations is how to use AI well. How can we gain productivity without losing quality? Using AI well means creating realistic guidance on when AI can support work, when human thinking must come first, how outputs should be reviewed, and which communications require clear human ownership.
Marc’s research ultimately shows that AI is changing work and revealing people at the same time: their needs, shortcuts, anxieties, creativity, loneliness, curiosity, and desire for support.
For communications leaders, the challenge is to embrace AI’s usefulness while protecting the qualities that make communication meaningful: judgment, care, originality, responsibility, and trust.