The experience of AI

Sketchnote showing how people interact directly with AI through chat, voice and clickable UI, and how they interact indirectly through AI-drive processes, people supported by AI and personalisation.

AI is not just a technology, it also creates an experience. So for organisations going through an AI transformation, simply adding AI components to your products and services is not enough. You also need to consider the human experience of interacting with AI for your employees, customers and wider stakeholders.

Direct interactions

As workers and consumers, our direct interactions with AI products involve the use of prompts or instructions to explain the tasks and outcomes we require.

These can be provided through voice, using products like Alexa and Sira or they can be provided through the humble chat window, which has become the primary interface for ChatGPT and other generative AIs.

Both voice and chat provide the scope for people to use the technology to perform very tailored tasks, but users need to be prompt literate to get the most out of them (similar to the difference between a novice and a highly skilled Excel user).

On the other hand, AI products with a single use case (such as a simple photo editor) may not need the flexibility of a chat window and can be designed with more intuitive, clickable interfaces. This makes it easier for new users to quickly master the product, but limits the potential for more advanced features.

Another factor is the quality of the content being produced. This is especially true for generative AI, which needs to meet consumer expectations. If you keep getting irrelevant answers to your questions, you’ll lose confidence in the product.

Indirect interactions

Even when not using an AI product, there is every chance you’ll be indirectly interacting with AI in the wider products and services you use. Everything from back office processes to customer-facing staff can be guided and supported by AI technologies. So the quality of this guidance impacts the experience you have.

Poor customer service may be annoying, but a bigger concern is biases unintentionally built into the AI programming that result in bad decisions been made about you. Credit and mortgage approvals being a prime example where someone could miss out on a better rate had there been more human involvement.

Then there’s personalisation. Amazon, Netflix and Spotify wouldn’t be the businesses they’ve become without tailoring their products and content to individual users based on past behaviours. As consumers, we’ve become used to (and expect) things to be personalised. But there’s also a point where personalisation becomes too personal, cheesy or just plain annoying.

User-centred design principles still matter, even with AI

Too often AI is touted as the solution to a gnarly problem, without fully understanding the problem. But no matter how clever AI products and services are, user-centred design principles are still needed.

To truly reap the benefits of AI technologies you need to keep the human experience front of mind. This means it’s vital to map out the customer journey and wider context, do the user research, be clear on the business and user needs, ensure usability and accessibility and find the best ways to craft positive customer experiences.

Thanks for reading 🙂

Scroll to Top