People care about the benefits, not the technology - AI design pattern

People don’t care about LLMs. Transformers. MCPs. Agents. Reinforcement Learning. People care about why it benefits them. What problem does your solution solve for them? What value does it bring them? Yes, you will likely be taking advantage of these technologies. And sure, you might need to put in all the fancy jargon for investors or your board, but for your end users? Just stick with how it helps them. 

Why do we need AI design patterns?

AI experiences are different from normal digital experiences. They are non-deterministic. They change based on many scenarios. People don’t trust what they don’t understand. They don’t want their software to seem like it is rolling dice. As a result, the data shows people don’t trust AI systems. This post is in a series on AI design patterns that is meant to help you improve outcomes and adoption of your AI experiences.

Understanding the ‘benefits over technology’ AI design pattern

Defining the design pattern

Too many teams get caught up in hype cycles and the technology that drives them. I saw this same pattern in the last round of AI/ML hype as well as the multiple iterations of Blockchain / Crypto / Web3 (crypto has never overcome this pattern). 

Teams want to share why their platform is using the latest and greatest tech. Detail on their landing pages what fancy models power their experiences. The problem? The average user doesn’t understand. And more importantly, they don’t care.

I am in the trenches of AI all day, every day. I even have a hard time telling the difference between o1, o3-high-mini, 4o, 4.5, Sonnet 3.7, Grok 3, DeepSeek R1, LlAma 3, Mistral Large, etc. Do you think the average user will understand?

They have a problem. They want to solve that problem. Simple. 

Does your banking application share if it is using Postgres, SQL Server or MongoDb? No.

Does your favorite social network advertise if it was written in Python, Ruby or PHP? No.

You could give a damn! You want to be able to check your balance and pay your credit card. You want to be able to catch up on the news or what your friends are up to in the world.

For some reason when it comes to new technology teams want to brag about the latest tech even though it is just a tool in the toolkit.

Aligned with human-centered design principles

This principle takes from good human-centered design patterns. You want to design solutions that solve real human needs. You want them to be easy to use. You want people to feel delight when using your product. 

Your goal when designing is to figure out how to get your user (who is a human) to value as fast as possible. How you accomplish that goal doesn’t matter ultimately.

AI design pattern: “People care about the benefits, not the technology”

Don’t even think about technology when you are designing your solution. That is something for you to work with engineering on, but it should not come into your designs. If you see a model named on the site that is user facing (and your user is not a developer) slash it! Share instead why your user should care about your solution. 

The key principles behind this pattern include:

• Clearly explain the benefit in words your users can understand
• Avoid technical jargon
• Identify the needs and outcomes and share those

Real-world examples of this pattern in action

Apple Watch

The Apple Watch uses AI to detect irregular heart rhythms or sleep patterns. If the watch spots something, it alerts users of potential health risks before it becomes a problem. Does Apple advertise that they use neural networks? No! They emphasize outcomes as saving lives and preventing hospitalizations before they occur!

Instagram / Facebook

For a long time social media companies have been using advanced machine learning to serve more interesting information and ads to you. It is not perfect, but still, I see more relevant ads on Instagram than I do watching sports on TV. These platforms have never broadcasted to users that they are using sophisticated algorithms, instead they focus on making their ads as relevant as possible. And giving users the ability to tell them that an ad is not useful to them (which makes their network smarter over time).

How to apply the principle in AI product design

Identifying user needs and desired outcomes

First, you need to understand what benefits you are solving. You can use workshops like our identifying & prioritizing AI use cases or do user research. You might already know what you are solving. If so, write that down too.

In particular, think about what your technology enables your users to do better.

Maybe your OCR and AI agent system helps your users save time filling in tedious forms. 

Simple sharing of benefits

Next up is creating visuals and text that clearly explains these benefits. Share the pain points or outcomes that you identified in the step before. List them out. Explain them. If you can share it in words that even a five year old can understand, even better!

Want help writing these clearly? Try plugging them into ChatGPT or Claud and giving them instructions to write a simple and clear benefit statement that a middle schooler will understand. 

Do this for each benefit or pain point individually. Make it simple.

Avoiding Technical Jargon

Tech jargon is confusing in general. It is especially bad with AI. People don’t get it. And in some cases it makes them trust your systems even less! If you have any reference to technology, try cutting it out entirely. Does it make a difference? Probably not. And if you are unsure, try testing your concepts for trust. 

AI can be so powerful. It is going to revolutionize how we live and work. Focus on that! Who cares about the models? Who cares about neural networks or deep learning? Just like people don’t care about your tech stack, they don’t care about your AI stack. We build software for humans after all, so make that AI software human-centered by being as clear as possible. And watch magic happen as users adopt and love your products!