5 Predictions to Watch for AI Design in 2024

2023 was an explosion of enthusiasm and experiments utilising Generative AI. Right now is one of the most dynamic times to build products, and lots of people are experimenting and putting stuff out. Some of it will work, much of it will not. What seems to be apparent though, even at this early stage, is that the how and what of design is undergoing a transformation as the application layer of Generative AI starts to get built out.

‍

From my vantage point, working at the intersection of Generative AI and User Experience every day for the past year, here are the things I’m looking at on the horizon. The throughline is that in 2024 we’re going to see more developed AI-native products launch and clear trends emerge within AI-enabled design workflows.

‍

My five predictions for 2024

‍

1. Everything Starts Getting Personal

This year, we will see the first wave of experiments utilizing dynamic customization of user interfaces and experiences based on individual user behaviors, preferences, and contexts. All enabled by AI. 

‍

AI-driven personalization will tailor the experience and the interface to individual user preferences, learning from user interactions to offer a unique experience. Eventually, each of us will have a 1-1 relationship with brands, products, and AI agents interacting with us and one another. These relationships are going to be a game changer, but also may have some downsides that need to be thoughtfully considered. 

‍

Some products already do this pretty well: Spotify’s Discover Weekly, Netflix’s recommendation engine, Amazon’s personalized product recommendations, Nike’s Run Club App, and Duolingo’s adaptive learning algorithm to name a few. Now imagine that capability coming down in cost to a fraction of what those companies have spent and deployed across every type of product and service you use. 

‍

Prediction:

• In 2024, personalization as a key part of the user experience will become much more accessible to all kinds of products. Where the barrier to entry for the companies doing it well was quite high; e.g., hiring, inventing new methods and processes, and deploying at scale…these barriers are quickly coming down with the advancements in LLMs and services to access them.
• A handful of startups will begin offering this type of personalization as a service for consumer facing brands, and it might be pretty good.

‍

These technologies will enable interfaces that adapt in real-time to each user's behavior, creating a more intuitive and engaging interaction. Generative AI will play a pivotal role in crafting these personalized experiences by autonomously generating design elements and content that resonate with individual user preferences.

‍

‍

2. Conversational Interfaces Actually Become Useful

It seems like voice interfaces have been ready to pop for the past 10 years, but it really does feel that way now with the step-change brought about with LLMs and their remarkable ability to parse human conversation. 

‍

The level of quality is generations ahead of what we’ve had the past few years which may signal to product makers it’s time to take this more seriously. The home and customer support seem like two very good candidates for wider adoption of this technology.

‍

We’ve already seen this kick off with the Humane PIN AI and the Rabbit r1, which are both predominantly voice interfaces. Throw Siri, Google Assistant, and Alexa in the trash. The new AI voice interfaces are smarter and easier to interact with. 

‍

I consider it likely that all three of them have been busily working on the next versions of these voice assistants because the home is such a rich opportunity for voice. But we will likely see voice appear in new applications, not just hardware, and perhaps even operating systems like Windows.

‍

Prediction:

• Relaunch of Alexa, Siri, and Google Assistant with greatly improved capabilities enabled by AI.
• At least one more new hardware form factor with voice as the main interaction mode. My bet is on earbuds.
• Voice moves to the laptop and phone with the ability to open and interact with a handful of applications, but this is just the beginning.

‍

‍

3. Teams Begin Adopting Generative AI Product Design Tools

The use of AI to create design elements in real-time, speeding up the design process and offering novel aesthetics goes mainstream this year.

‍

Design tools have always been at the cutting edge. First, with Photoshop bringing graphic design to the computer and then Macromedia’s suite of applications custom built for a then new medium of web design. The last ten years have seen innovation in digital prototyping products like Sketch and Figma which support digital product and interaction designers.

‍

It will continue now with a new generation of AI-native design products like Semanttic, RunwayML, Vizcom, Midjourney, and others. The promise is that AI will enable designers to automate some of their work so they can focus on higher value activities. With each new generation of new tools, the tools shape the creators. The conclusion is we have entered a new era of design. The era of designing AI experiences and products with AI-native tools. 

‍

Prediction:

• Expect to see the most innovation from startups, like Semanttic, in this area. Without the sunk cost of years (or decades) into existing products and the manual mode of work, they are the most likely to innovate this new AI-driven workflow from first principles.
• Expect to see the established players (Figma, Adobe, all of the “low-code/no-code” builder tools) release AI powered features that try to reduce incremental points of friction.
• We will see more and more designers experiment and prototype new workflows by stitching together many different AI tools as this area evolves.

‍

Bonus Prediction:

• Prompt Engineering becomes Prompt Design, and more designers start moving upstream to design the prompts. I see this akin to when designers started moving upstream in the web 1.0 era and information architecture was born. Prompt Design is a new type of information architecture.

‍

‍

4. The User Journey Becomes Central

As more and more of the UI work gets offloaded to AI, designers will begin to have the tools and the time to advocate for, and work across, a journey based approach.

‍

In 2024, AI's improving capability to predict user paths and provide intuitive navigation means that we will start to see new tools that enable anyone working across a user journey to reduce friction and enhance user engagement in ethical and valuable ways.

‍

Prediction:

• We will see some early experiments with customized navigation paths. AI will start to dynamically adjust the navigation structure or suggest the most relevant next steps.  The purpose will be to create a more personalized and less overwhelming experience, but it also serves as an important interim step to when all interactions are AI mediated, real-time, and instantly composable. 
• Better context-aware assistance. As a reasoning engine, AI can offer much better context-specific guidance and support than the decision trees of today. For instance, if a user seems to be struggling with a particular task in a part of the journey, the AI can proactively offer help or suggestions with the proper context. No more unhelpful chatbots that walk you through a decision-tree and then dump you into the top of an FAQ. In 2024, we will start to see this in ecommerce first.
• Seamless Omnichannel Experience. We will see some early signals of this, and in 2025 this will start to become much more prevalent. It’s the holy grail of surfacing up the exact right information and options at the exact right time across any type of screen and location. AI will enable a cohesive experience across different platforms and devices, remembering user preferences and actions from one platform and applying them to another. Shopping and traveling seem to be two standout areas where this could fundamentally change today’s user experience.

‍

These advancements will not only make digital interfaces more intuitive and user-friendly but also foster deeper engagement by making users feel understood and valued by the systems they interact with. Part of the challenge is technical, but there’s a potentially even greater challenge in user adoption. What’s coming is a wholly new way of interacting with digital systems from what we have today. We have all developed a lot of habits, modes of thought, and ways of working that will not apply in this new era.

‍

‍

5.  Every Website Becomes an API

We founded Semanttic, in part, because we clearly see a future where the user interface isn’t just built by AI but is also interacted with by AI on our behalf. Rabbit just released the r1, a pocket device running their LAM (Large Action Model) which promises to do just that. By talking to this device, you can book your Uber or order food from Door Dash. It works not through APIs but by using the same website and application user interfaces humans use.

‍

In this new era, every website will become an API. As AI becomes more adept at interacting with web interfaces in human-like ways, it's also plausible that we will start optimizing websites for AI interactions. This could lead to designing interfaces that are more standardized and AI-friendly, emphasizing clarity and predictability in layout and navigation. It suggests a paradigm where the traditional boundaries between front-end interfaces and back-end APIs blur. Websites themselves could serve as 'de facto' APIs, accessible not just to humans but also to AI agents. After all, why invest resources in building an API when AI can abstract what it needs from the human user interface.

‍

Prediction:

• Diminishing Need for Traditional APIs. If AI can effectively navigate and interact with web user interfaces, the necessity for dedicated APIs will reduce in the medium to long term, especially for simpler tasks. This could lead to a shift in how data is accessed and manipulated, with AI agents acting as intermediaries. But my prediction for 2024 is we will see a few very high-traffic services (e.g., Uber) experiment with this in addition to continued support for their existing APIs.

‍

• Accessibility and Standardization Play Key Roles. Somewhat ironically, the tools and support we have been building the past years to make our products more accessible to diverse groups of people can also benefit AI utilization. With accessibility now “dual-purpose”, there may be greater funding for accessibility initiatives, which could have positive ramifications for human users.

‍

For all of us working with Generative AI in Design, it's important to consider how this shift will impact the user experience for humans. The primary focus of all design should remain on human users, with AI compatibility being an additional, but not overriding, consideration in my opinion.

‍

Finally, 2024 is an opportunity to lean even more into user-centric design practices, where understanding and responding to user needs is not just ideal but essential. Not every product will benefit from the predictions above. But this new Generative AI era will enable a compelling pathway for products that are smarter and more intuitive, and user journeys that feel friendly and personal.

‍

If you have found these predictions interesting, please follow me on LinkedIn or X and check out what I’m building at Semanttic. Let’s go 2024!

‍