Embracing human-centred design in AI: A path to ethical and beneficial innovation

As a user researcher with experience in UK government digital services, Google, and neuroscience, I've always been fascinated by the intersection of technology and human needs. The rapid advancement of artificial intelligence (AI) presents both exciting opportunities and significant challenges for society.

Recently, I've been exploring the critical role that human-centred design (HCD) plays in shaping AI systems that are not just technologically advanced but also ethical, beneficial, and aligned with human values. This exploration is based on the paper titled 'Human-centred AI: The role of Human-centred Design Research in the development of AI,' authored by Jan Auernhammer. The paper was presented at the DRS International Conference 2020, held online from August 11th to 14th, 2020, and published by the Design Research Society in the conference proceedings for DRS2020 - Synergy. The paper is available on the DRS Digital Library.

The limitations of a purely rationalistic approach

Traditionally, AI development has been dominated by what researchers call a "rationalistic perspective". This approach views humans as "cognitive machines" and focuses primarily on replicating human abilities through mathematical and technological advancements. While this has led to impressive technical achievements, it often falls short in addressing the complex realities of human experience and societal impact.

A purely rationalistic approach can:

Oversimplify human behaviour: Humans are not simply rational entities operating on logic alone. Emotions, cultural contexts, social influences, and subjective experiences all shape how we perceive and interact with the world. Viewing humans as cognitive machines can lead to AI that lacks empathy or fails to understand the nuances of human needs.

Neglect important ethical considerations: Reducing humans to computational systems often brings a narrow focus on technical optimisation and efficiency. Ethical dilemmas, such as fairness, justice, and the impact on well-being, may be overlooked, particularly when efficiency is prioritised above human values.

Struggle to anticipate unintended consequences: A focus on replicating cognitive abilities without considering human context can lead to unintended consequences. For instance, AI may solve a particular problem efficiently but cause unforeseen social or psychological impacts.

Fail to meet real user needs and experiences: Designing AI based solely on cognitive replication might lead to systems that are highly capable but lack intuitiveness or accessibility. By focusing too much on mimicking human cognition, AI systems may fail to align with real-world user needs or be unintuitive to interact with.

The power of human-centred design in AI

This is where human-centred design comes in. By placing human needs, experiences, and values at the forefront of AI development, we can create systems that are not only intelligent but also beneficial and ethical. Here are some key HCD approaches that are crucial for the future of AI:

Human-centred systems: Examining how AI impacts social structures and organisations.

Social design: Considering the designer's responsibility and the societal consequences of AI.

Participatory design: Involving diverse stakeholders in the design process.

Inclusive design: Ensuring AI systems are accessible and useful for as many people as possible.

Interaction design: Shaping meaningful and intuitive human-AI interactions.

Persuasive technology: Understanding and ethically managing the persuasive potential of AI.

Human-centred computing: Investigating how AI integrates into everyday life.

Need-design response: Addressing genuine human needs responsibly through AI.

A call for collaboration

A rationalistic perspective treats AI as an intellectual pursuit to replicate human cognitive abilities. However, this view can overlook human complexity. Humans are shaped by emotions, ethics, social interactions, and experiences.

For AI to benefit humanity, we need a balanced approach that combines rationalistic advances with design thinking. This means considering the social, ethical, and cultural dynamics that shape our lives.

AI should not only think like humans but also enhance human experiences, respect dignity, and adapt to societal complexity. A human-centred approach ensures AI is beneficial, ethical, and aligned with human needs.

By embracing these principles, we can create AI that is not only intelligent but wise—AI that learns and enhances human life.

Developing human-centred AI requires collaboration across disciplines. Designers, technologists, policymakers, and others must work together to enhance human capabilities while respecting societal values.

In my work, I advocate for and implement these human-centred approaches. Whether through user research, co-design sessions, or advising on ethical issues, I believe putting humans at the centre of AI is crucial.

As AI shapes our world, let's ensure it amplifies human potential, respects diversity, and contributes positively to society. The future of AI is about creating a smarter, more empathetic, and human-centred world.