Generative AI: Rapid Uptake and Major Benefits

Generative AI usage in the UK has seen rapid growth – almost 40% of adults were using it by mid-2024, and over a quarter of employed people were incorporating it into their work routines. This pace of adoption is faster than that of PCs or the internet, with productivity gains across tasks like writing, coding, and data analysis.

Most current AI interactions still take place through basic chat interfaces, but this is just the beginning. There is significant potential in designing smarter, more tailored AI systems that can offer richer, more specific experiences beyond the basics.

‍

What is Conversational AI?

Conversational AI enables natural, human-like interactions with AI tools – think of chatbots that genuinely understand what you’re saying, with the ability to interpret speech, text, and even different languages. These tools make AI feel accessible, allowing more people to use it daily in an intuitive way.

The key value isn’t just in the intelligence of the AI but in how easy it is to interact with. Tools like ChatGPT have brought AI to the public by making conversations with it feel simple and natural. The aim now is to design AI that’s not only smart but also feels human, reflects a brand’s tone, and truly helps people.

‍

The Language and Experience of Conversational AI

When we talk to an AI, every word matters. The way an AI phrases things impacts the experience: language should be clear, friendly, and appropriate to the user, whether they’re customers or employees.

This is where NLU (Natural Language Understanding) comes in. With NLU, AI can grasp the context of what people are asking, not just their exact words, enabling more flexible, meaningful interactions. It helps AI handle unexpected questions or phrases, making the interaction feel smoother and more “human.”

‍

Designing AI for Multimodal Interactions

We’re moving towards an era where users will interact with AI in ways beyond just text or voice. Multimodal AI can recognise gestures, facial expressions, and other cues, creating richer, more dynamic interactions. This opens up opportunities for businesses and designers to craft experiences that feel intuitive and seamless.

However, building these types of interactions comes with its own challenges. Sometimes, AI may mistakenly associate unrelated elements in the data, like always linking a baby with a dummy in every context. Managing these errors will be crucial as AI becomes more advanced.

‍

How Timing Affects Trust in AI Interactions

Interestingly, slight pauses in AI responses can actually increase engagement, as people interpret these delays as the AI “thinking.” These brief pauses give users a moment to process, making the eventual response feel more thoughtful. Simple prompts like, “Let me find that for you,” turn latency into a design feature rather than a flaw.

‍

Real-World Examples: DoorDash & Klarna’s Use of Conversational AI

DoorDash and Klarna are both using conversational AI to improve support interactions. DoorDash, for example, employs a system that pulls relevant information for users and double-checks for accuracy before sending responses. Klarna keeps conversations helpful by using specific “guardrails,” ensuring the AI doesn’t stray off-topic. For more complex cases, a human agent steps in to maintain a smooth user experience.

‍

The Future of Conversational AI: Memory and Personalisation

Traditionally, AI systems start each conversation afresh, forgetting any previous interactions. But that’s changing. Companies like Convergence are building AI tools that remember past interactions and learn from them, creating experiences that are more personal and seamless. OpenAI is also exploring memory in its systems, adding voice capabilities and making AI feel like a familiar, helpful assistant.

Memory is set to play a significant role in Conversational AI, transforming it from a transactional tool to something that genuinely supports businesses and individuals in smarter, more efficient ways.

So, if you’re looking to design or implement conversational AI, remember: creating a natural, intuitive experience is key. Whether it’s incorporating memory, handling latency gracefully, or ensuring accuracy, the future of AI lies in crafting smart, human-like interactions that bring real value to users.