G’day!

​

In the rush to embrace AI, there is a concerning trend amongst some companies cutting their UX research budgets and teams. Its so short-sighted.
​
​

​
​The true power of AI lies in creating a synergistic partnership between humans and machines; not replacing one with the other. This is where UX research becomes not just valuable, but indispensable.

​

So lets dive in shall we...

Why UXR matters more than ever?

​

​

​

1. AI Doesn’t Understand Humans Automatically: AI is great with data. But it doesn’t get human feelings or cultural nuances. I think we can all agree—what is online is not completely representative of the real world or your life within it. And we all know how often user testing comes back showing users behaving in ways we didn’t expect. UX researchers are like translators between AI and humans.
   ​
2. Trust is Hard to Build, Easy to Lose: For people to use AI, they need to trust it. Right now there is a huge trust gap that exists between the average user and AI systems. UX research helps figure out how to build that trust.
   ​
3. One Size Doesn’t Fit All: What works in one country might flop in another. UX researchers help make sure AI works well for different people, places and contexts. As hyper-personalisation rises, we must share insights about users’ many personal contexts with dev teams. This will help them create valuable personalisation without being “creepy.”
   ​
4. Ethical AI Needs Human Input: AI is only as good as the data you feed it and the mental models of development teams building the algorithms. Humans don’t just crunch numbers. Our choices spring from a moral compass forged through years of living. This ethical framework, not raw data alone, guides our decisions. We filter information through a lens shaped by our experiences, values, and beliefs. What is just, what is right and what is fair. It’s an internal algorithm that spreadsheets full of data can’t match. AI systems without human ethics will make unfair decisions. Like a little kid, they won’t know any better until you teach them.

​

AI reshapes UX research, transforming traditional methods and researcher roles. Real-world cases reveal the critical nature of user experience studies. As technology evolves, so do the skills and approaches needed in this field. UX professionals must adapt swiftly, embracing new tools while preserving core principles. This shift demands a delicate balance between innovation and established practices.

UX Research 2.0: Evolving for the AI Age

The field of UX research must continue to transform to meet the challenges of AI development. Here’s some key ways that needs to happen:
​
​Human-AI Collaboration: The Cornerstone of Successful AI Integration
​
​AI systems must understand human experiences and mental models. This is key to building reliable, productive partnerships with users. This core principle underpins the development of truly collaborative artificial intelligence.
​

Consider an AI that quickly processes data but misses human decision nuances. Or a chatbot that replies fast yet lacks empathy in delicate situations. These examples highlight the need for Human-AI Collaboration in AI development. They also show why UX research is vital.

​

Why Human-AI Collaboration Matters
​
​Trust Building: For users to rely on AI systems, they need to trust them. UX research finds what drives trust in different contexts. It also shows how to design AI interactions that foster it.

​

Leveraging Complementary Strengths: Humans and AI have different strengths. UX research helps find where AI can boost human skills. It also shows where human oversight is vital.

​

User adoption: Neglecting human factors can render sophisticated features irrelevant. Engagement stems from addressing user needs, habits, and preferences. Without this focus, even groundbreaking AI may languish, untapped and overlooked. Success demands a holistic approach. It must blend tech skills with a deep understanding of human behavior and motivations.

​

Ethical Considerations: Human-AI collaboration raises complex ethical questions. UX researchers can uniquely investigate and address these issues. They will do so from a human-centric view.

​

Case Study: Grab’s AI-Powered Driver-Partner System
​

Grab’s approach to creating a driver-friendly AI system. It offers lessons in human-AI collaboration.

​

​
​

Contextual Inquiry: Researchers accompanied drivers to observe their work and difficulties. This method uncovered challenges not visible in the data.

​

Co-design Workshops: Grab ran workshops for drivers, designers, and AI engineers. They collaborated to design the driver app’s interface and features.

​

Behavioural Pattern Analysis: Researchers studied driver behaviour and decision-making. This helped create an AI that complemented human intuition, not replaced it.

​

Feedback Loops: Grab set up a system for drivers to give immediate feedback on AI suggestions. This created a continuous learning loop for the AI.

​

Cultural Adaptation: The research found big differences in driver preferences and behaviors across Southeast Asia. This led to local versions of the AI system.

​

​

The result? An AI system that boosted driver efficiency and satisfaction. Now, 90% of drivers find the AI suggestions helpful in their work.

​

​

Conquering the Qualitative Data Mountain

Qualitative research has always faced a limit. We can only gather, analyse, and report on so much data in a set time. Gone are the days of drowning in interview transcripts, desperately searching for patterns. AI is changing the game:

​

Automated Transcription and Tagging: Tools like Dovetail are changing our approach to qualitative data. AI transcription cuts manual work significantly. Meanwhile, smart tagging quickly identifies themes in many interviews.

​

​

Theme Clustering at Scale: AI algorithms can now analyse vast amounts of qualitative data. They can find patterns and clusters that might escape even the most eagle-eyed researcher.

​

Privacy and Compliance Boost: Dovetail is also using AI to help us comply with regulations like GDPR. It automatically finds and removes sensitive information.

​

​

Triangulating Research: The North Star

Using multiple research methods and data sources has always been the gold standard. But, it’s become more important now. AI is data-hungry. It’s vital for models to learn to interact with a wide range of users. That means diverse data is required to ensure systems are fair and minimise bias. Let’s look at how Spotify approached this with their Discover Weekly feature:

​

Case Study: Spotify’s Discover Weekly

Spotify’s Discover Weekly, launched in 2015, is a personalized playlist. It uses AI to recommend music to users.

Here’s what we know about its development:

​

Quantitative Data Analysis: Spotify mined vast troves of user data to fuel their AI. Listening patterns and behaviour from millions shaped the algorithm. This deep dive into user habits informed the qualitative research plan.

​

Qualitative Research: Ethnographic studies yielded diverse user insights, helping refine the feature. While exact methods remain undisclosed, this research played a key role in shaping the final product. The company’s commitment to human-centric design shone through this comprehensive feedback-gathering process.

​

Feedback Loops: Spotify fine-tunes its algorithm constantly, driven by user interactions and insights. This process keeps refining the platform to match listeners’ tastes.

​

The result: Spotify’s Discover Weekly gained over 40 million users in its first year.

​

​

​

Beyond Self-Reporting: The Biometrics Revolution

Self-reported data’s flaws have long hindered UX research progress. Users often struggle to accurately convey their experiences, leading to skewed insights. Researchers grapple with this unreliability, seeking alternative methods to capture genuine user behavior and preferences. AI-powered biometric tools are opening new doors:

​

Eye Tracking: AI can analyze eye-tracking data. It can reveal patterns in users’ visual attention that they may not know.

​

Voice Analysis: It can find insights that traditional methods might miss. AI-powered voice analysis detects subtle changes in tone and emotion.

​

Brain wave analysis via EEGs: It reveals users’ hidden reactions. It offers insights into split-second interactions with AI interfaces. EEG taps into the subconscious, revealing subtle responses traditional testing might miss.

This window into neural activity is valuable for evaluating fast digital experiences or micro-interactions. Researchers can use it to focus on questions about split-second decisions or feelings.

​

I have personally used all three of these biometrics in combination with user testing of our product experience vs our competitors. Being able to show the different in the amount of frustration, joy and cognitive load for users throughout different parts of our experience were so impactful.

Ethical AI: From Concept to Practice

​

As AI systems grow more powerful, we must align them with human values. For UX researchers, this means developing new methods to study human-AI interaction. For developers, it’s about creating flexible AI systems that can adapt to human input. For product managers, it’s about fostering a product culture. It should value both human expertise and AI capabilities.

​

​

Google’s Model Cards project illustrates this evolving role:

Transparency in AI means all users must understand how it makes decisions. They should be able to explain what data AI is using and interpret its decision-making process or actions taken.
​
Transparency is vital for several reasons: It helps build trust between AI systems and their users, it enables us to hold AI systems and their creators accountable for decisions and outcomes. Being able to understand AI decision making allows us to identify and correct errors, biases, or unintended consequences in AI systems.

​

Here is an example of how Google researchers helped build the framework for transparency of their AI models.

​

Google introduced Model Cards in 2018 as a framework for transparent reporting of AI model information. Here’s what we know about their development:

​

1. Google forged a collaborative path in its design process. UX researchers engaged AI developers, product managers, and outside partners. This diverse input shaped the team’s approach, blending perspectives to craft a well-rounded solution.

​

2. Iterative Design: They refined the Model Card format via user tests. This ensured users could understand the key details and make informed decisions.

​

3. Focus on Transparency: The project aimed to improve communication about AI models’ performance and limits. It targeted both technical and non-technical audiences.

​

The Result: Google and other tech companies now use Model Cards as the industry standard. They improve transparency in AI development.

​

​

New Frontiers in Researching Human-AI Collaboration

​

As UX researchers, we must develop new methods to study and design for human-AI collaboration. Here are a few:

​

Collaboration Journey Mapping: Create detailed maps of users’ and AI systems’ interactions over time. Identify key moments of synergy and friction.

​

AI Partnership Heuristics: Create design principles for effective human-AI partnerships. Focus on trust, control, and mutual understanding.

​

Hybrid Intelligence Usability Testing: Design tests to evaluate not just the AI or the human user in isolation. Also, test how well they work together as a unit.

​

AI Transparency Studies: Examine how AI transparency affects user trust and collaboration. Consider different levels and types of transparency.

​

Long-term Collaboration Impact Assessment: Conduct longitudinal studies to understand how human-AI partnerships evolve over time and their long-term effects on user behaviour and skills.

​

​

💡 Key Takeaway for Product Teams: It’s not just about fancy AI algorithms. It’s about a productive partnership between humans and AI. For UX researchers, this means developing new methods to study human-AI interaction. For developers, it’s about creating flexible AI systems that can adapt to human input. For product managers, it’s about creating a product culture. It should value human expertise as well as AI capabilities.

The Consequences of Neglecting the Human Element

​

Here’s the stark reality: neglecting UX research in AI development can lead to significant issues. The potential costs include:

​

Failed Adoption: Advanced AI features may go unused if they don’t meet human needs and expectations.

​

Erosion of Trust: Badly designed AI can quickly erode user trust. This can lead to resistance to future AI projects.

​

Missed Opportunities for Innovation: Companies may miss unique ways AI can boost users. They need to understand human behaviour’s nuances to avoid this.

​

Ethical Missteps: Without user research, AI systems may perpetuate biases. They may also make decisions that users find unethical or unfair.

​

​

As we integrate AI into our products, remember this: The goal isn’t to replace human intelligence. It’s to enhance and empower human abilities with AI. This is the essence of human-AI collaboration. It’s why UX research isn’t just important—it’s essential in the AI era.

​

​

Why This Matters for Everyone

You might be thinking, “This is cool, but I’m not a UX researcher. Why should I care?” Well, here’s why it matters for different roles:

​

Product Managers: UX research helps you make better decisions about what to build. It can also help build trust in AI systems. This may boost adoption or save you from wasting resources on unwanted features.

​

UX Designers: UX research gives you real data to back up your design choices. It helps you create designs that not only look good but actually solve user problems.

​

Developers: Understanding user needs helps you build better AI models. It can also help you spot potential ethical issues early on.

​

Business Leaders: Good UX research leads to happier customers. That usually means better business results.

​

Wrapping Up

​

As AI becomes a bigger part of our lives, UX research becomes more important, not less. It’s the bridge between smart machines and human needs. It helps create AI that’s not just clever, but truly helpful and trustworthy.

​

Remember, the goal is to create AI that makes human lives better. And that’s exactly what good UX research helps do.

​

Next time an AI app seems to “just get you,” remember: there’s likely great UX research behind that magic!

​