← Back to main blog

The next stage for Design Thinking towards Design with Data + AI

Ethereal hands form a luminous circle of collaboration, blending golden warmth with cool teal in a flowing dance.
Photo by Stockcake

Over the past decade, design thinking has been a go-to methodology for innovation: empathy, ideation, iteration. But in a world of complexity, interdependence, scale, and speed it’s starting to show its limits.

When challenges are tightly constrained, human-centered processes shift and adapt well. But when problems are emergent, uncertain, and volatile those same linear cycles often struggle to keep pace.

We also need to recognise that human behaviour is inherently difficult to predict. Systems built on rigid, binary thinking or overly simplified algorithms will almost certainly fall short.

A well-known example comes from the U.S. housing crisis: some companies used mathematical models to estimate home values and assess lending risk, assuming people would behave in stable, rational ways. But human behaviour didn’t follow the models. Banks approved loans for people who couldn’t realistically afford them, and when the market shifted, many homeowners simply walked away. The algorithms failed because they treated a complex human system as if it were predictable and linear and, the real world isn’t.

But that will be the material for a new blog post.

So yes, Design thinking revolutionised problem solving by centering empathy, creativity, and iteration, a needed counterbalance to purely analytical methods. But the world has evolved. Many of today’s challenges, from global supply chains to personalisation at scale, are too dynamic, data-rich, and interconnected, and on the ubiquitous age a purely human-centered, linear design process to handle effectively.

That’s where data and AI come in. AI helps us sense patterns, model complex systems, and generate insights or solutions built on top of data and human intuition. In a sense, we are moving from Design Thinking to “design sentience”, blending human empathy with computational foresight.

So while design thinking isn’t “dead,” it’s evolving. The next frontier is augmented sentience: using AI and data not to replace human design, but to expand what’s possible, seeing relationships, predicting outcomes, and testing ideas at a scale and speed never before possible.

Instead of asking, “What do people need?” we might also ask, “What do the data reveal about unmet needs we can’t yet articulate?” The synergy of human imagination, intuition and experience with machine intelligence is the new design paradigm.

I think the next frontier is augmented sentience: combining human insight with data, AI, and systems thinking.

Aspect Deterministic AI Generative AI
Definition Produces a fixed, predictable output for a given input, the same input always yields the same result. Produces new, variable outputs for the same input, responses are generated probabilistically.
Example (Simple) A calculator: 5 + 5 always equals 10. A language model asked “Write a poem about the ocean” will create many possible poems, each slightly different.
Underlying Mechanism Rules-based, algorithmic, or model-driven systems that follow explicit logic trees or trained decision boundaries. Deep learning models (especially large language or diffusion models) that sample from learned probability distributions to “create” text, images, code, etc.
Behavior Predictable, consistent, explainable. Creative, flexible, context-sensitive but less predictable.
Best For Tasks requiring precision, repeatability, and verifiability, e.g. accounting, data retrieval, rule-based automation. Tasks requiring creativity, synthesis, or language understanding, e.g. writing, design ideation, conversation, generative design.
Drawbacks Limited adaptability, can’t handle ambiguity or open-ended inputs. Can “hallucinate” (generate plausible but incorrect outputs), lacks guaranteed accuracy.
AI Examples
  • Predictive analytics model that forecasts demand using set parameters.
  • Traditional rule-based chatbot (“if user says X → reply Y”).
  • ChatGPT, Midjourney, DALL·E, Claude, Gemini, etc.
  • Generative design tools that create new layouts, marketing copy, or art.
Analogy Like following a recipe: exact steps → exact result. Like a chef improvising: draws from experience, tastes, and creativity to make something new each time.

Main conclusions

The interplay between deterministic and generative responses is becoming critical for creating effective products.

Increasingly as systems become more asynchronous and distributed, modern products, particularly those powered by AI depend on deterministic and generative processes operating simultaneously.

Deterministic systems provide reliability, safety, consistency, and predictable outcomes.
Examples: calculations, validation checks, compliance rules, payment flows.

Generative systems provide creativity, flexibility, adaptation, and contextual understanding.
Examples: recommendations, natural-language interfaces, personalised content.

Why the communication between the two matters?

Products now often require a blend of:

Examples of this interplay in real products:

So yes, the relationship and balance between deterministic and generative components is becoming a key design principle in modern product development.

Reading list

Books and papers related to systems thinking, ripple effects, and design & technology.

← Back to blog