Skip to main content
    Insights · Climate Intelligence

    Designing human-centered AI.

    How IDEO's philosophy shapes the way Lightspark builds climate intelligence — an essay on data, trust, and the design of action.

    By James Riley7 min readMay 2026

    For decades, IDEO has influenced how the world approaches innovation. Their philosophy of design thinking reframed design from something associated primarily with products and aesthetics into something much larger: a methodology for solving complex human problems.

    IDEO's central idea is deceptively simple. The best innovations emerge when organizations deeply understand people — their behaviours, frustrations, motivations, fears, and aspirations — and then use technology thoughtfully to improve their experience.

    DesirablehumanFeasibletechResponsibleethicsViablebusinessInnovation
    Fig. 01The DVFR framework — desirable, viable, feasible, responsible. Innovation lives where all four overlap.

    This philosophy has shaped everything from consumer technology to healthcare, education, public policy, and urban systems. Increasingly, it is also shaping how artificial intelligence products are designed.

    At Lightspark, this way of thinking fundamentally informs how we approach climate intelligence, housing data, energy modeling, and risk systems.

    Because while AI is powerful, technology alone does not solve human problems. Thoughtful design does.

    I

    The problem with most technical systems

    Historically, many enterprise systems have been designed around institutions rather than human beings. Climate systems are a perfect example, and often the homeowner is the least prioritized in the design.

    Homeowners encounter a fragmented maze of retrofit programs, energy audits, contractor quotes, financing options, insurance requirements, rebate systems, and technical reports they do not fully understand.

    Similarly, financial institutions and insurers face their own complexity. Climate risk data, property characteristics, emissions information, catastrophe models, and underwriting systems frequently exist in disconnected silos.

    The result is friction everywhere: confusion, disengagement, low adoption, reactive decision-making, and missed opportunities.

    "What experience is the user actually having?"

    That question is deceptively powerful because it changes where innovation begins. Instead of starting with technology, design thinking starts with empathy.

    "A human-centered approach to innovation that draws from the designer's toolkit to integrate the needs of people, the possibilities of technology, and the requirements for business success."

    — IDEO
    II

    Designing AI around human outcomes

    At Lightspark, we do not view AI as the product itself. We view AI as an intelligence layer that helps people make better decisions. That distinction matters enormously.

    A homeowner is not searching for a probabilistic flood-return curve or an energy simulation model. They are trying to answer far more personal questions:

    • Will my home be safe in the future?
    • What upgrades should I prioritize?
    • How much will they cost — and will they reduce my bills?
    • Will they protect the value of my property?

    Banks and insurers are not simply looking for more datasets. They are trying to understand where risk exists, how climate exposure changes over time, and how they can proactively help customers future-proof their homes before catastrophic loss occurs.

    These are not purely technical questions. They are behavioural, financial, emotional, and operational — simultaneously.

    Fig. 02The homeowner's emotional arc — moving from anxious uncertainty to confident resilience.
    "The challenge is not merely producing more intelligence. It is designing systems that make intelligence understandable, actionable, and trusted."
    III

    One window into property intelligence

    One of the core concepts behind Lightspark is the idea of creating one window into property intelligence. This idea emerged from observing how fragmented the housing and climate ecosystem has become.

    Today, information about a single property may exist across dozens of disconnected systems — climate hazard databases, municipal records, utility information, insurance systems, energy models, geospatial datasets, retrofit assessments, and financing tools.

    Fig. 03Lightspark sits at the center of the housing ecosystem — a unified layer connecting risk, energy, finance, and action.

    Most people — including sophisticated institutions — cannot meaningfully navigate that complexity. Design thinking encourages organizations to step back and rethink the experience from the user's perspective.

    Instead of asking

    How do we expose more data?

    The better question

    How do we reduce complexity while increasing confidence and action?

    That shift in perspective changes the entire design philosophy. At Lightspark, this has led to systems that combine climate risk, energy performance, carbon modeling, retrofit pathways, resilience scoring, and upgrade economics into unified workflows.

    Fig. 04From hazard signal to reduced loss — the path data must travel to become outcome.
    IV

    Why explainability matters in AI

    As AI systems become more sophisticated, human-centered design becomes even more important. The future of AI will not belong solely to organizations with the largest models or the most data. It will belong to organizations capable of building trust.

    In climate and housing, trust is critical because the decisions involved are deeply personal and financially significant. Homes represent safety, identity, long-term wealth, and security. That means AI outputs cannot feel abstract or opaque.

    People need to understand why a risk score exists, what factors influence it, what actions can improve outcomes, and what the economic implications may be over time.

    Fig. 05The Lightspark stack reads bottom-up — raw signals climb through modeling and explainability into clear human outcomes.

    Good design reduces cognitive burden. The best systems make complexity feel intuitive without oversimplifying reality. The goal is not technical accuracy alone — it is confidence, clarity, and engagement.

    V

    From static software to adaptive systems

    Innovation emerges through iteration rather than perfection. Traditional enterprise software was often designed as static infrastructure: define requirements, build once, deploy broadly.

    But AI systems behave differently. They evolve continuously alongside user behaviour, new datasets, changing climate conditions, and shifting regulations.

    Fig. 06Two diamonds, four moves — discover and define the problem space; develop and deliver the solution. Repeat indefinitely.

    At Lightspark, product development is approached as a continuous learning system: prototype rapidly, observe behaviour, refine workflows, and improve continuously.

    VI

    Climate adaptation is a design problem

    Many societal challenges are not purely technological. They are systems-design challenges. Climate adaptation is one of them.

    The world already possesses enormous amounts of climate science, engineering knowledge, and building technology. Yet adoption remains slow because the surrounding systems are difficult to navigate.

    The challenge is designing better experiences, better incentives, better workflows, better engagement, and better pathways to action.

    "The future of climate intelligence depends on combining advanced data science with thoughtful systems design."
    Lightspark

    In many ways, this is precisely what IDEO envisioned decades ago: technology designed not around complexity itself, but around the human beings trying to navigate it.

    End
    — fin —Essay 04 / 12
    Share this essay