Stanford CS547 HCI Seminar | Spring 2026 | Observing the User Experience in 2026

Key Concepts

• Ground Truth: The core, empirical reality of human experience that must be verified through direct interaction, as opposed to synthetic or AI-generated data.
• 3E Framework: A lens for understanding why knowledge work is being automated: Extraneous (seen as secondary), Expensive (costly to hire specialists), and External (outsourced or automated).
• Zero-Trust UX: A methodology for participant vetting that assumes potential fraud (AI-generated personas) and requires continuous verification.
• PWDR (Person Who Does Research): A term acknowledging that research is now a shared organizational competency rather than a siloed job title.
• Organizational Power: The ability to influence decisions, which is often tied to the perceived authenticity and "social contract" of research deliverables.

---

1. The Evolution of UX Research in the AI Era

The speakers, authors of Observing the User Experience, argue that the discipline has shifted from a "how-to" focus to a "why" focus. The rise of AI has fundamentally changed the practice:

• Automation of Tasks: AI has commoditized repetitive tasks like transcription, coding, and drafting discussion guides.
• The "Tidal Wave" Effect: AI is not a single event but a continuous force mashing up existing practices, making research faster but also more prone to "anti-aliased" reality—data that looks correct but has the "corners rounded off."
• The Fraud Problem: AI has enabled "hyperscaled fraud," where participants use LLMs to fake expertise. The speakers advocate for zero-trust vetting: delaying payments, asking anomaly-probing questions, and verifying identities continuously.

2. Synthetic vs. Real Research

• Synthetic Users/Personas: Useful for identifying basic requirements or reaching groups that are difficult to recruit. However, they are inherently limited by their training data (the "corpus").
• The "Walled-Off" Problem: LLMs are trained on public internet data. They fail to capture the reality of professionals working behind firewalls (e.g., healthcare policy experts) or populations whose lives are not documented in English-language digital spaces.
• The Need for Ground Truth: Traditional, in-person research remains essential for external validation, identifying edge cases, and collecting stories that AI cannot synthesize.

3. Organizational Power and the "3E" Framework

The speakers emphasize that AI is a symptom of a broader shift in power dynamics, not the sole cause of industry changes.

• The 3E Framework: Organizations automate work they deem Extraneous (not core to the product), Expensive (requires specialized headcount), and External (can be offloaded to machines).
• The Role of the Researcher: To avoid being labeled "extraneous," researchers must understand their organization’s internal power structure. They should use the same research skills to understand their stakeholders (e.g., what makes a VP successful) as they do to understand users.
• Deliverables as Social Contracts: A research report is not just a document; it is a contract between the researcher and the stakeholder. If a report is AI-generated, it lacks the "aura of authenticity" and the social weight of a human having done the work, which reduces its influence within the organization.

4. Practical Strategies for Professionals

• Speak the Language: Do not force organizational stakeholders to learn UX terminology. Use their vocabulary and format (e.g., creating PowerPoint slides that are easily "stolen" and reused by others).
• Proof of Work: Use "hacks" to establish credibility. For example, taking photos of yourself in the field (wearing a bright shirt) to prove you were physically present and engaged with the user.
• Embrace Friction: The speakers argue that resistance from communities (like tribal nations) regarding data extraction is a healthy, necessary friction. It forces organizations to be more ethical and intentional about their research practices.

5. Synthesis and Conclusion

The main takeaway is that while AI tools can accelerate the process of research, they cannot replace the value of research. The "ground truth" is like a raw material—it must be mined and refined through human effort. As the industry moves toward a model where research is a shared competency (PWDRs), the professional's role is to act as a bridge between empirical human reality and organizational decision-making. The speakers conclude that if you are "inventing the future," you cannot rely on a corpus built from the past; you must go out and talk to real people.