Using AI well means learning how to prompt

AI products are fundamentally changing how users interact with software by making the quality of output dependent on the skill level of user prompts. Unlike traditional software where mastery leads to consistent results for all skilled users, AI tools create a collaborative experience where the same prompt can yield different yet equally valid outcomes based on nuanced user intent and context. This shift presents unique challenges for product teams who must find ways to guide users toward successful outcomes despite varying prompt expertise.

The AI experience dilemma: AI products differ fundamentally from traditional software because the quality of output largely depends on the user’s prompting ability.

• As Loïc Houssier, VP Product at Superhuman, noted, “Now with LLMs, a bunch of the perceived quality depends on your prompt… the outcome may be perceived as low quality, but that’s something that is really hard to control.”
• While mastering traditional software like Photoshop leads to consistent results among skilled users, AI tools create a spectrum of valid outcomes based on nuanced intent and context.

Product management approaches: Companies are developing strategies to improve user experiences despite varying levels of prompt expertise.

• Many product teams are rewriting user prompts behind the scenes to expand on user intent and transform basic queries into more nuanced, successful responses.
• Even with prompt rewriting, anticipating how users might want to steer AI interactions remains challenging for developers.

Follow-up questions as a solution: Implementing clarifying questions creates a more natural and effective AI interaction pattern.

• ChatGPT exemplifies this approach by asking follow-up questions to refine broad queries, mimicking how a human colleague might seek clarification.
• These questions serve a dual purpose: they help the AI gather necessary information while simultaneously helping users better understand and articulate their own requests.