In the fast-moving world of generative AI, prompts have become the gateway to innovation, productivity, and creative exploration. But how people use prompts — and how those usage patterns shift — offers even deeper insights into human-AI interaction.

The recent Forbes article, What We Do in the Prompts — OpenAI, Anthropic and Ipsos on AI Usage by Diana Spehar, highlights emerging trends in prompt behavior and adoption. When viewed through the lens of Nate Patel, the discussion becomes not just observational but strategic, uncovering lessons for developers, enterprises, and policymakers alike.

Key Insights from the Forbes Article

The article explores how prompt usage is evolving across platforms like OpenAI and Anthropic, and how data firms such as Ipsos analyze adoption patterns. Some of the key points include:

• Shifts in prompt types — from casual exploration to problem-solving and task automation.
• External factors shaping behavior — including pricing changes, model updates, and platform policies.
• Challenges in measurement — due to sampling biases and gaps in data visibility.
• Opportunities in prompt analytics — helping companies align models with real user needs.

These insights set the stage for Nate Patel’s interpretation of what matters most.

Nate Patel’s Perspective on AI Usage Patterns

1. Prompt Types as Behavioral Signals

Nate Patel emphasizes that analyzing what people ask is more meaningful than raw usage numbers. A rise in prompts asking for refinement, iteration, or clarification suggests users are pushing AI systems toward higher productivity — not just experimentation.

2. External Shocks Drive Real Change

He cautions against attributing usage dips or spikes to seasonal cycles alone. Instead, Patel points to external shocks — like pricing shifts or policy updates — as the primary drivers of sudden behavioral change.

3. Measuring Signal vs. Noise

For Patel, platform usage data is only one piece of the puzzle. True insights require triangulating data from multiple sources — logs, feedback, and third-party analytics — to avoid misleading narratives.

4. Prompt Analytics as Strategy

Rather than treating analytics as passive observation, Patel frames it as a strategic lever. Prompt data can shape product roadmaps, improve user experience, and even guide future AI capabilities.

5. No One-Size-Fits-All Interpretation

Patel highlights that each AI provider’s ecosystem is unique. Usage patterns on OpenAI may not mirror those on Anthropic or others. Overgeneralizing, he warns, risks missing crucial context.

Explore the insights: Nate Patel on How We Shape AI through Our Questions