Design Thinking is Not New, But Essential for Human-Centered AI

Firdaus Adib
Published in
4 min readSep 15, 2023


Image Association of MBA

Design thinking has existed for decades, providing a human-centric approach to developing products and services. While not a novel concept, design thinking offers an important perspective as artificial intelligence continues rapid advancement.

Historically, AI development has been driven by technology, missing the core essence of problem-solving and understanding user needs. However, using design thinking, in the beginning, to define problems and make models may take longer, but it can speed up progress by finding mistakes early and avoiding wasted work later in development. Design thinking can help re-orient AI development around human needs.

Refocusing AI on Meeting Actual User Needs

AI used to be about technology and making it better, but to really solve problems, we need to understand what people want. By using design thinking, we can make AI development more focused on what users need.

Design thinking orients the development process around those needs, providing an essential counterbalance to technology-driven AI. While new methods aren’t needed, this mindset shift is key.

Diagram by Cue Logic

Accelerating Progress By Identifying Pitfalls Early

The iterative approach of design thinking requires more time upfront spent framing problems and prototyping concepts. It’s slower at first, but it reduces the need to redo work later on. For complex AI initiatives, avoiding fixable mistakes can greatly accelerate impact.

Balancing Subjectivity With Objectivity

Subjectivity refers to personal biases, opinions, and interpretations that are influenced by individual experiences and perspectives. In design thinking, subjectivity can play a role in the ideation and prototyping stages, where designers may be influenced by their own preferences and assumptions.

On the other hand, objectivity refers to a more impartial and unbiased approach to problem-solving. In AI, objectivity is essential in ensuring that algorithms and models are not influenced by personal biases or prejudices. This is important to prevent discriminatory outcomes and ensure fairness in decision-making.

Design thinking brings inherent subjectivity through its emphasis on empathy and diverse user perspectives. This needs balancing with software engineering rigor and objective technical requirements.

Combining design thinking with robust technical practices introduces needed oversight without losing a human-centric viewpoint. Integrating these contrasting approaches is challenging but essential for ensuring AI is both creative and effective while also being fair and unbiased.

Integrating Into Tech Teams, Even In Small Ways

Cross-functional teams are ideal for design thinking, but can be challenging in technical AI projects dominated by engineers.

Even if the project is complicated, starting with small steps like talking to users and making simple models can help a lot. When using design thinking, you look at all parts of the problem, like the data, the models, and how people will use the product. This helps you find new ways to solve the problem. If the project affects people directly, it’s important to have people with different skills to make sure everyone is considered.

Having a Product Manager who understands both product and technical aspects like AI really helps.

Aligning Solutions With Product Roadmaps

To achieve a complete integration of design thinking when implementing artificial intelligence (AI) initiatives, it is crucial to ensure that these initiatives are aligned with the broader product roadmaps and objectives. This alignment is necessary to guarantee that the solutions developed through design thinking are centred on humans and contribute to the attainment of the overarching goals of the product.

To connect user experience (UX) improvements with business outcomes, it is essential to use techniques like impact mapping. This technique helps to establish the relationship between UX improvements and their impact on business outcomes. Furthermore, prioritization frameworks, such as RICE, are also useful in sequencing AI prototypes for efficiency. By using RICE, you can prioritize AI initiatives based on their potential impact, confidence level, and resources required, which ensures that you achieve the best possible results with the available resources.

Dovetailing with Agile Sprints

Day-to-day, design thinking workflows should complement agile sprints. Discovery sprints frame problems, design sprints prototype concepts, and development sprints build minimum viable AI. Regular user testing synchronizes with sprint reviews. This cadence integrates design and development, but flexibility ensures possibilities aren’t constrained.

We should test the product regularly while working on it. This will help designers and developers work together smoothly. But we should also be flexible and not limit ourselves too much.

In Summary

Using just design thinking is not enough, but if we combine it carefully with software and data science best practices, we can make the most of its benefits. It helps us keep a perspective that focuses on people so we can balance the fast progress of AI. If we integrate design thinking and AI thoughtfully, they can work together to create solutions that can bring big changes.



Firdaus Adib
Editor for

Web craft. Rails. Data Science. Biohack. Currently learning iOS.