Designing for a New AI World

The Shift to the Intent-Based Design

For the last 60 years, we have mostly interacted with computers in a command-based paradigm. We provide a direct input to get a direct result and continue this back-and-forth feedback loop until we reach the desired outcome. The introduction of AI has greatly influenced the way we interact with computing and has caused a shift to an intent-based paradigm. Instead of using a set of commands, we now express our intent to the system and then collaborate with it by refining inputs and guiding AI as it interprets, adjusts, and responds dynamically.

With such a drastic change in human-machine communication, it is obvious that previously established approaches to design will not work anymore. Designing intent-based systems calls for new ideas, fresh perspectives, and unconventional solutions to meet emerging computing needs, interfaces, and experiences.

A more intuitive and flexible interaction with systems made them human-like for end-users. This newfound anthropomorphism in computing demands we consider a human behavior and relationship standpoint in our design methodologies and approaches.

The way people articulate intent is inherently human – messy, contextual, and full of ambiguity. You can definitely remember the times your words were misinterpreted or misunderstood beyond your intention. The same kind of ambiguity exists in intent-based AI systems, and we need to learn to counter it.

Here, we can refer to our behavior science expertise to design effective human-computer relationships. As we start to perceive computing systems as our AI counterparts, we can account for that and create better ways of working with them by incorporating human relationship models, and cognitive biases, and heuristics into our designs. Understanding how an AI system can build trust with humans over time is one of the examples of how we apply these frameworks.

Designing for a World of Intelligent, Invisible Systems

We may also be on the edge of a paradigm shift in computing modality. The emerging partnership between Johny Ive and Sam Altman hints at a future that moves away from the screen altogether. Has Mark Weiser’s ubiquitous computing finally become reachable? If anything, we are definitely taking our first steps in this direction. More and more computing will be taking place in the background without human intervention, driven by intelligent systems.

However, before we reach this level of computer autonomy, we need to find answers to essential design questions: when should interactions come to the surface and require a human touch? How do we know that the systems are following our intent? What level of affordance is the right level? These questions require extensive research, more testing, experimentation, and continuous iteration to try to form the answers.

At the heart of this shift is systems thinking and system design. We can no longer design products in isolation but must keep in mind the complexity of a multifaceted system that both AI and humans interact with, leaving just enough on the surface for efficient communication. Considering all relevant factors and applying systems thinking skills and tools are now critical to success in this design work.

Take the concept of Mobility 2.0, for example. Imagine a future of Software-Defined Vehicles (SDV) – autonomous co-pilots that take care of your every need and want. Your only input is the destination. The question is how to design such a seamlessly working system. Beside anticipating the needs of the driver, the car must be in tune with every other system that compile the smart city infrastructure – other vehicles, the roads themselves, pedestrian safety AIs, energy grids, public transportation systems, and so on. All these components are powered by their own AIs to make them safer, more efficient, and more autonomous. To be able to fit their singular product into a multilayer ecosystem, designers must be aware of it as a whole and apply systems thinking to creating a harmonious part of a consistent experience.

The Critical Role of Data in Human-Centered AI

Finally, we cannot talk about designing AI systems without mentioning data. Designers are often trying to make sense out of the vast data available – collected in the deep research, acquired from clients, derived from AI. Unfortunately, on any step of its processing, data can be colored with unintentional bias, impacting the end result. As designers, we must be deeply aware of how our own processes of gathering, structuring, and handling data influence what is used and how.

At the same time, data offers unprecedented opportunities for personalization that we could have only imagined before. The power of it, however, must be kept in balance with ethical considerations of responsible AI to decide what’s right for the user and how we continue to provide for human agency in these interactions.

Conclusion

To sum up, AI is causing major shifts in the things we need to consider and implement as designers and businesses. Designers need to reimagine their skillsets to include behavior science, systems thinking, and data science and apply them accordingly to preserve our humanity as we move forward with this new technology. The future is here, and we finally have the chance to build systems we have imagined for years. The potential for us to untether form screens and embrace the era of quiet computing is here. The question now is: will we design it in a way that centers people, preserves agency, and strengthens the human experience?

Learn more about our approach to AI