.jpeg)
The rapid expansion of AI is not just creating a pool of opportunities but also empowering users to address their pain points. Thanks to AI, we can now facilitate information finding and make complex tasks more accessible and straightforward, instilling a sense of confidence and control in users.
I've been in the technology industry for about seven years. For the last two years, I've worked almost exclusively on creating new AI products or adding AI features to products from different sectors, such as investments, cybersecurity, travel, and legal. While I don't think AI will kill design, adapting our design work to suit this latest wave of technological innovation is essential.
Here are the design patterns I've been using and testing. They became something essential to consider when I'm starting to design an AI-driven product:
1. Provide quick replies and show system status.
In an ideal world, all users should be able to get a quick reply, but we know that currently is technologically impossible. However, we can always inform the user about the system status and what the AI is doing in the backend, making them feel valued and integral to the process.
2. Guide the users to cut costs and boost revenue.
As product designers, we must remember that some AI interactions can elevate stakeholders' costs, resulting in less revenue for your clients. For example, when designing a chatbot, every question the user creates will elevate the price. You can make suggestions to guide the users through interactions to solve this. This way, you are not limiting their interaction but guiding them in how to use the product.
Consider whether it is better to provide multiple options in the interface and allow the user to select and customize to get the expected outcome instead of a chatbot.
3. Allow the user to give feedback and get powerful insights.
AI development is still far from where users expect, so there will always be a possibility that the AI will not give a correct response.Always give users the option to send feedback when this happens. Their feedback is not just important; it's crucial. Ensure your product saves all this information in the backend; this helps improve the AI, the product, and, eventually, the user experience, making them feel heard and integral to the process.
4. Design for error prevention.
Users can make mistakes, so we always provide a way to recover from and prevent them. But what happens when the AI makes a mistake?When the AI makes a mistake, the only one who will know it is a mistake is the user, so the user will have to return from that error. That's why we always need to display a way to regenerate or re-do the action so the user can try again; we need to ensure it is as easy as possible.
When defining the product, we can identify some possible technical errors that the AI can make; we can prevent them by informing the user.
5. Provide transparency in the process.
Communicating how the AI works is not just ornamental; it's crucial to building and maintaining user trust in the AI-driven product.Good communication involves displaying the AI's process understandably and transparently. Some users, especially those with complex processes, must trust a completely automatic process; some will be unhappy. To address this, always showcase sources, calculations, and the location where the AI made that finding. Providing a way for the user to review and accept the AI suggestions is an excellent way to automate processes but not take away the user's trust.For instance, you can show the data sources the AI used to make a recommendation or the calculations it performed to arrive at a decision.
As designers, it's crucial and our responsibility to understand and adapt to the design patterns behind AI-driven products. These patterns are not just a trend but an inevitable part of our future. Users are already expecting them, so we must familiarize ourselves with them and take a proactive role in their implementation.
