This feature interview traces eight years of research by Professor Zhong-Ren Peng of the University of Florida on artificial intelligence and urban planning. It follows his path from the four-stage typology of urban planning AI proposed in the “Pathway” paper to Symbiotic Planning Theory and the CORE framework for governed human-AI cocreation. Throughout the conversation, one concern stays central: the capacity-legitimacy tension in planning. This is the question of how planning can expand its analytical capacity without weakening democratic legitimacy. The interview also focuses on the “compliance trap,” where following formal procedures can coexist with real inequity. Professor Peng argues that building accountability into planning infrastructure, and developing comparative methods across cases, will be the next frontier for making AI-enabled planning research cumulative and testable.
Editorial Introduction
Artificial intelligence is rapidly reshaping how cities are studied, governed, and planned. Yet AI’s importance for planning is not just about faster analysis or stronger optimization. It raises a deeper question: how can planning expand its analytical capacity without weakening democratic legitimacy, public accountability, and equity? This feature interview traces Professor Zhong-Ren Peng’s eight-year engagement with this question, from a 2018 early conference paper on AI and urban planning to two journal articles that mark major milestones in the field. The first was published in the Journal of Planning Education and Research (Peng et al., 2024), referred to below as the “Pathway paper.” It describes a development path from AI-assisted planning to AI-autonomized planning, and clarifies how the planner’s role might change as machine capability expands. The second was published in the Journal of the American Planning Association (Peng et al., 2026), referred to below as the “Symbiotic Planning Theory paper” or “SPT paper.” It proposes Symbiotic Planning Theory and the CORE framework as a governance-centered paradigm for human-AI cocreation in urban planning, illustrated through the Gainesville micromobility case and its “compliance trap.” Together, the two papers complete a shift from describing how AI enters planning to considering how AI’s operation within planning should be governed, and this interview points further ahead toward future research directions. To explore these questions further, the editorial team at GUIHUA: Frontiers of Urban and Regional Planning interviewed Professor Zhong-Ren Peng. The interview transcript follows below.
Article title
From pathway to symbiosis: rethinking urban planning in the age of AI
Keywords
Symbiotic Planning Theory, Human–AI co-creation, Algorithmic accountability, Equity governance, Urban Planning AI, Micromobility
Authors
Zhong-Ren Peng, Xiaoxiao Feng & Zhiqiang Siegfried Wu
Journal
Frontiers of Urban and Rural Planning, 2026, 4:13. https://doi.org/10.1007/s44243-026-00086-5
I. Where the Idea of Urban Planning AI Began
GUIHUA:FURP Editorial Team:
Your research on AI and urban planning spans eight years now, from a 2018 conference paper to two major journal articles since. What is the core question that runs through this whole research path?
Professor Zhong-Ren Peng:
The core question is what I call the capacity-legitimacy tension in planning. How can planning strengthen its analytical capacity without eroding its democratic legitimacy? Cities today are asked to make more decisions in less time and under deeper uncertainty. Climate risk, housing pressure, transportation inequity, and cumulative environmental burdens all demand faster decisions that also carry technical credibility. At the same time, the scale of urban data now far exceeds what a planning team can meaningfully analyze on its own. So the question is not just whether AI can help. The real question is whether planning can use AI to expand its capacity to observe, simulate, and test options, without letting technical systems quietly replace public judgment.
This concern became clearer to me after living through the smart city era. The original vision of the smart city was understandable. It promised better data, better models, and better urban management. But in practice, many smart city projects did not produce more democratic governance. They produced governance that was harder to see through. Important decisions about transportation, investment, and risk became embedded in technical systems that communities could neither review nor question. That lesson has stayed with me.
My 2018 conference paper, published in Urban Planning Forum, was my first attempt to lay out this problem clearly for the Chinese planning community. The later “Pathway” paper responded to it by asking how AI enters planning and what new roles it might take on. But over time, I became increasingly convinced that this field needs more than a typology of technical change. It needs a governance and methodological paradigm, one that can explain how human planners, communities, and AI systems work together without breaking down democratic accountability. My keynote at the 2025 AAAI “AI for Urban Planning” workshop reinforced this view, and that line of thinking still shapes my research today.
GUIHUA:FURP Editorial Team:
“Urban AI” is now a widely used term, but you draw a clear line between “urban AI” and “urban planning AI.” Why does this distinction matter so much? What does it mean for how we think about urban governance and accountability?
Professor Zhong-Ren Peng:
This distinction matters because not every use of AI in cities counts as planning in the institutional sense. “Urban AI” is a broader category. I see it as an analytical framework within urban science, built around three core functions: identifying patterns, analyzing causes, and optimizing urban systems. It includes methods like machine learning, deep learning, and computer vision, as well as AI-enabled methods for causal analysis. These get applied to things like travel demand prediction, land use change, environmental monitoring, and infrastructure management.
“Urban planning AI” is a specific subfield of urban AI. It places these analytical capabilities inside the institutional context of democratic plan making. Specifically, it refers to using AI agents, in part or in whole, within the planning process itself, to support, augment, automate, or make autonomous certain parts of plan preparation and policy formation. Once AI enters this space, it no longer just helps us understand cities. It participates in shaping public decisions about land use, transportation, infrastructure, zoning, and investment.
This difference matters because planning is not just a technical activity. It involves public power, competing values, institutional procedures, and accountability to communities. When we use AI to analyze a city, we are doing science. When we use AI to allocate resources or set zoning for a community, we are exercising public power. In that second case, the question is not just whether the model is accurate. The question becomes whose goals get served, which constraints actually bind the system, who can challenge it, and who is accountable for the final decision.
So this distinction is useful conceptually, but it also matters politically. It lets us pin down what responsibilities planners and public agencies still hold once AI becomes part of the planning process itself. It also explains why governance becomes the central issue the moment AI moves from urban analysis into public decision making.
GUIHUA:FURP Editorial Team:
In your “Pathway” paper, you laid out a development path for urban planning AI. What do you see as that paper’s main contribution? What background and concerns led you to write it?
Professor Zhong-Ren Peng:
That paper became necessary because discussion of AI in planning was growing fast, but the field lacked a clear vocabulary and a coherent map of how things were developing. We looked into this through a scoping review and found that the literature was expanding quickly while staying highly fragmented. As of May 2023, a Web of Science search combining “Artificial Intelligence” and “Urban Planning” returned 881 papers, and 74 percent of them were published after 2020. That showed clearly that the field had strong momentum and needed conceptual clarity urgently.
The paper’s main contribution was to define “urban planning AI” more clearly and to propose a four stage pathway: AI assisted planning, AI augmented planning, AI automated planning, and AI autonomized planning. We were not trying to predict that every part of planning would inevitably move through all four stages. Instead, we wanted to give the field a framework for understanding how AI’s role might shift, gradually, from supporting plans to actually making them. Related to this, the paper also clarified a possible shift in the planner’s role, moving from “planner in the loop” to “planner above the loop,” and in some limited cases, potentially toward “planner out of the loop.”
I think the value of that paper is that it moved the conversation beyond isolated tools and case studies. It showed that AI is not simply another technical upgrade to existing planning support systems. It has already started reshaping the structure of the planning process itself. At the same time, the paper was careful to point out that this shift raises serious questions about fairness, transparency, trust, and the planner’s continued role. In that sense, the “Pathway” paper was written intentionally as foundational work. It mapped out the problem, clarified the trajectory, and raised a question it did not yet answer itself: not just how AI enters planning, but how AI’s growing role should be governed so it does not erode democratic accountability. That question became the central driver behind the Symbiotic Planning Theory paper.
II. From Pathway to Symbiotic Planning Theory
GUIHUA:FURP Editorial Team:
The “Pathway” paper mapped a path from AI assisted planning to AI autonomized planning, but a pathway is not the same as a governance theory. What questions did that typology leave unresolved? And what ideas or experiences pushed you toward Symbiotic Planning Theory?
Professor Zhong-Ren Peng:
The “Pathway” paper gave the field a way to describe how AI enters planning, but it did not answer the harder question of how AI’s growing role should be governed. A typology can distinguish different stages of autonomy and different shifts in the planner’s role, but it cannot answer the toughest normative questions on its own. When should AI propose a solution, and when must a human authorize it? How do public values get turned into binding constraints? What kind of accountability is needed as machine involvement deepens? In that sense, the paper was meant as foundational research, but it was not complete.
I became increasingly aware that once AI starts moving from supporting plans to making them, the core issue is no longer just technical capacity. It becomes institutional legitimacy. AI can expand the space of possible solutions, reveal patterns humans might miss, and compress work that once took months into a matter of days. But if these capabilities are not governed, what looks like analytical progress can quietly turn into a shift of power. That power can move into systems that communities, and even decision makers themselves, can no longer truly review or question.
What sharpened this concern was the research that later became the Gainesville micromobility case. We observed a pattern where formal compliance with equity rules coexisted with real, deep inequity on the ground. That turned out to be the deeper lesson. The problem was not just an imperfect model or an imperfect policy. Governance arrangements can meet their own formal standards while still failing the very communities they claim to protect.
That realization pushed my thinking beyond typology and toward theory. It also led me to a question the “Pathway” paper had not asked directly: not just how AI enters planning, but how we design institutional conditions so that delegated analytical power can still be recalled and held accountable by the public over time.
GUIHUA:FURP Editorial Team:
Why do you think a new planning theory, Symbiotic Planning Theory, was necessary?
Professor Zhong-Ren Peng:
A new theory became necessary because existing planning paradigms were not built for a world where AI can generate, simulate, and refine large numbers of alternative plans at machine speed. Rational planning, participatory planning, and codesign have all contributed something important to the field. Rational planning gave us analytical rigor. Participatory planning gave us democratic legitimacy. Codesign gave us iterative creativity. But all three assumed that human cognition still set the practical ceiling on planning capacity. AI has changed that assumption.
This brings us back to the tension that has driven this research from the start, the capacity-legitimacy tension. AI expands planning capacity by letting us process more data, test more scenarios, and spot patterns or relationships we could not see before. But it also strengthens a real risk. Important choices can migrate into technical systems whose internal logic is hard to review, and whose authority can become normalized simply through everyday use. That is why I came to believe this field needs more than ethical guidelines or better tools. It needs a paradigm that explains how AI and human planners should relate to each other institutionally.
Symbiotic Planning Theory is my answer to that need. Its central claim is that AI should be treated neither as a passive tool nor as an autonomous decision maker. It should be treated as a governed cocreator that operates under binding human direction. This middle position matters a lot. If we treat AI only as a tool, we underestimate its power to shape knowledge production, problem framing, and scenario generation. If we treat it as a decision maker, we risk eroding the democratic foundation of planning. Symbiotic Planning Theory tries to hold both truths at once. It protects democratic authority over planning goals, while using AI to expand and test the range of feasible means.
So the issue is not just that planning needs a one time update to adapt to AI. The deeper issue is that AI has changed the structure of planning itself. Once that shift happens, theory has to keep up.
GUIHUA:FURP Editorial Team:
Symbiotic Planning Theory rests on three necessary and sufficient conditions, and it depends on a mechanism you call “governed friction.” Can you explain what these concepts mean? Why do they add up to a genuinely new paradigm, rather than a repackaging of existing ideas?
Professor Zhong-Ren Peng:
What makes Symbiotic Planning Theory distinct is that it defines “symbiosis” institutionally rather than rhetorically. It is not simply saying that humans and AI should work together. Many planning tools already do that in a loose sense. Symbiotic planning is more specific. It treats AI as a cocreative partner, but its generative power must operate under binding governance, clear processes for contestation, and human authorization that can be revoked. Without these conditions, what we have is still just augmentation, not symbiosis.
The first condition is normative triggering. AI must surface the normative questions embedded in a task. It has to turn vague commitments like “we want this to be fair” into explicit, formal constraints that can actually govern the process. Value judgments should not get quietly buried inside an objective function or a set of parameters.
The second condition is binding agency. Community priorities cannot stay at the level of advisory input. They have to become hard constraints the model must satisfy. This is exactly what separates symbiosis from consultation. If stakeholders establish an equity requirement, the model must be retrained to meet that requirement before it can be deployed again.
The third condition is time bound primacy. AI’s “license to operate” expires unless it is renewed through evidence based human reauthorization. This matters a great deal, because it prevents algorithmic authority from becoming normalized through institutional inertia. Every reauthorization cycle forces a clear human judgment: does this system still serve the community’s evolving values and priorities? It works both as a check against drift and as an opportunity for institutional learning.
These three conditions get operationalized through what I call governed friction. The core idea is this: what actually makes this process democratic is not seamless automation, but productive tension. The mechanism unfolds in three steps. Generative provocation is when AI produces an efficiency driven or simplified baseline plan, which exposes hidden tradeoffs. Normative recalibration is when planners and communities turn those tensions into binding constraints. Binding reauthorization is when the model gets retrained, and its continued use depends on monitored outcomes, not just procedural compliance.
What makes this a new paradigm is precisely what earlier traditions could not offer. Rational planning never explained how machine generated alternatives should be governed. Participatory planning never explained how public values become hard algorithmic constraints. Codesign brought iteration and creativity, but it usually treated technology as a facilitation tool rather than an active co-producer of knowledge. Symbiotic Planning Theory brings these threads together under the new conditions AI has created, and it adds what earlier paradigms lacked: explicit mechanisms for contestability, constraint formation, and time bound democratic control.
III. Operationalizing Symbiotic Planning: The CORE Framework and the Gainesville Case
GUIHUA:FURP Editorial Team:
How does the CORE framework turn Symbiotic Planning Theory into a workable planning process?
Professor Zhong-Ren Peng:
The CORE framework is an operational method that turns Symbiotic Planning Theory into a repeatable planning process. CORE stands for Collaboration, Options, Refinement, and Execution. Its core sequencing principle is simple but important: accountability comes before algorithms. In other words, the planning process does not start with optimization. It starts by defining goals, guardrails, equity floors, participation rights, and challenge procedures, before the model is even allowed to generate options.
During the Collaboration stage, planners, communities, and relevant agencies define the problem together and establish a social contract for AI’s involvement. This includes setting goals, identifying what counts as harm, clarifying participation rights, and deciding which constraints must bind the model. One key principle here: a gap in the data is treated not just as a technical problem, but as a possible equity signal that may call for targeted outreach.
During the Options stage, AI generates alternative solutions rather than a final answer. Through model pluralism and structured scenario generation, it can present competing pathways organized around different priorities, such as efficiency, equity, or resilience. The value of this stage is not that AI finds a single best solution. Its value is that it stops value judgments from getting quietly buried inside the objective function. It makes tradeoffs visible, and open to democratic challenge.
During the Refinement stage, these options get reviewed, questioned, revised, and rejected when necessary. This is the stage where community priorities move from advisory preferences to binding constraints. Humans hold absolute veto power. Planners record why certain proposals get accepted, modified, or rejected, which creates an audit trail for normative judgment.
During the Execution stage, implementation is treated as the start of monitored governance, not the end of planning. The model’s continued use depends on outcome based performance, not just procedural compliance. Throughout the CORE framework, planners take on three distinct roles: guardian of judgment, conductor of AI, and keeper of ethics. These roles are supported by real governance tools, including model documentation cards, audit trails, KPI dashboards broken down by equity dimension, appeal pathways, and records of time bound reauthorization. These are not optional add ons. They are institutional infrastructure that turns implementation into a cycle of learning, correction, and accountability. This is how the CORE framework keeps AI acting as a governed cocreator, rather than an unaccountable source of authority.
GUIHUA:FURP Editorial Team:
The Gainesville micromobility case sits at the center of the Symbiotic Planning Theory paper. You describe a “compliance trap,” where the vendor met the formal deployment rules, but equity outcomes still fell short. What does this case reveal? Would these problems have been visible at all without the Symbiotic Planning Theory framework?
Professor Zhong-Ren Peng:
The Gainesville case shows that procedural compliance can mask real inequity. The vendor met the rule requiring at least 10 percent of scooters to be deployed in the equity zone. But that equity zone only produced about 3 percent of all rides, while about 84 percent of rides started near campus. This is a compliance paradox: following the rule can coexist safely with a failed outcome on the ground.
Symbiotic Planning Theory reframes this as a governance problem, not simply a weak enforcement problem. The existing rule focused on input compliance, meaning where the devices were placed, rather than on outcomes. The monitoring system it relied on had no mechanism to trigger an institutional response when the gap between the two grew wider.
The AI component clarified this diagnosis further. We first used a deep reinforcement learning model to generate an efficiency optimized baseline plan. That baseline concentrated deployment around the campus area, which essentially reproduced the existing pattern of service. We did not treat this output as an answer. We treated it as a generative provocation. It revealed what optimization produces when equity is not set as a constraint.
What came next was the most critical step, what we call normative recalibration. The city carried out systematic community engagement on this program for the first time in four years. Surveys and workshops revealed structural barriers that no dashboard alone could explain: gaps in walking accessibility, a mismatch with free public transit options, affordability concerns, and payment exclusion facing unbanked households. These concerns were translated into binding model constraints. In effect, the algorithm’s goal shifted from “maximize usage” to “maximize usage subject to an equity floor.”
So the SPT framework let us see the difference between a program that looks fair on paper and a program that actually runs fairly on the ground. Without the generative provocation that came from the deep reinforcement learning baseline, the compliance trap might have stayed invisible. Without translating structured community engagement into binding constraints, a technical fix would have just reproduced the same spatial inequity through a more sophisticated optimization process. Without outcome based reauthorization, even an improved deployment plan could drift back over time toward a campus centered pattern. This case shows that AI can be a powerful tool for revealing hidden inequity. But only a governed human-AI process can ensure that this revelation actually turns into accountable, lasting change.
GUIHUA:FURP Editorial Team:
Looking at both papers together, what do you see as the biggest challenge, risk, or limitation facing AI in planning?
Professor Zhong-Ren Peng:
I want to be honest about this, because I think the planning field sometimes swings between uncritical enthusiasm and reflexive suspicion, and neither one actually helps us. AI in planning is genuinely powerful, but its structural risks are just as powerful. It would be irresponsible of me to present SPT and the CORE framework without also acknowledging the limits of this approach itself.
The most widespread technical risk is bias rooted in data. AI systems learn from historical patterns, and historical patterns in urban data often encode decades of underinvestment, exclusion, and unequal service provision. When a model trained on vendor telemetry data recommends concentrating scooters near campus, it has not made a technical error. It is simply reproducing faithfully what the data shows. The problem is that “low demand” often reflects structural exclusion rather than genuine preference. An AI system that cannot tell these two things apart will reproduce inequity systematically, at machine speed and machine scale. Better algorithms alone cannot fix this. It requires a governance design that treats data gaps as equity signals and builds community knowledge into the modeling process as a binding input, not as an after the fact correction.
The second risk is opacity. Many AI systems, especially deep learning models, do not produce outputs that planners, let alone communities, can easily review or question. When a model recommends a zoning boundary or a transit route change, the reasoning behind it can genuinely be hard to reconstruct. This opacity creates a particular danger in planning, because public justification is not optional. A decision that cannot be explained cannot be legitimately challenged, and a decision that cannot be challenged cannot achieve democratic accountability. This is exactly why the CORE framework insists on treating model documentation cards, audit trails, and explainable drivers as governance infrastructure, not as technical extras.
The third risk is a capacity constraint, and I think this one is underappreciated in the academic literature. The vision of governed human-AI cocreation described in the SPT paper requires a fairly high level of institutional capacity. It requires staff who understand AI systems well enough to evaluate their outputs. It requires procurement frameworks that can guarantee audit rights and interoperability. It requires community engagement infrastructure that can turn lived experience into binding constraints. It requires monitoring systems connected to real governance consequences, not just passive reporting. Many planning agencies, especially smaller local governments and those serving communities that have been neglected for a long time, do not currently have this capacity. This is not a small logistical gap. If governed human-AI planning only works for well resourced agencies, the framework risks reproducing planning inequity rather than correcting it.
There is also a deeper structural risk. I believe both papers try to respond to it, but no single governance framework can fully solve it on its own: the risk of optimization replacing public judgment. This is what I call the “deployment legitimacy fallacy,” the dangerous assumption that legitimacy gained at launch will automatically persist. The danger is not only that an AI system might produce biased or opaque outputs. The danger is that efficiency itself can gradually become a substitute for legitimacy. A recommendation that looks technically defensible can quietly replace the harder political work of deliberation, contestation, and democratic authorization. Governing this risk requires more than better technical design. It requires a planning culture that always treats AI as a provocateur and a tool, never as an authority.
I should also be honest about the limits of my own current work. The SPT paper is theory building work. It was validated through a single city case, and it is limited by vendor data and non random community input, so its causal inference and transferability are limited. The next step has to be comparative and longitudinal research, testing the SPT mechanism across multiple cities, domains, and governance settings; that is the direction I plan to keep pursuing.
IV. From Pathway to Paradigm, and Future Directions
GUIHUA:FURP Editorial Team:
How do you now see the relationship between the “Pathway” paper and the Symbiotic Planning Theory paper? Are they a sequence, a correction, or something else?
Professor Zhong-Ren Peng:
I see them as two complementary stages of the same research program. As I said earlier, the “Pathway” paper was developmental and diagnostic. It gave the field a shared vocabulary for describing how AI’s role might evolve. The Symbiotic Planning Theory paper responded to the challenge that came next: once AI starts moving from supporting plans to making them, under what institutional conditions can that role still hold democratic legitimacy?
If the “Pathway” paper mapped the terrain, the SPT paper defined a governed paradigm within it, one that treats AI as a cocreative partner operating under binding human direction and reauthorization. The second paper is not a correction of the first. It deepens the problem the first paper had already revealed.
There is also an important methodological progression between the two papers. The “Pathway” paper was built on a scoping review, aiming to organize fragmented literature into a coherent framework. The SPT paper shifts toward combining theoretical synthesis with empirical demonstration. This progression reflects the development of a research program: from mapping what the literature shows, to proposing and testing a governed arrangement within the terrain that literature reveals.
Taken together, the two papers complete a move from pathway to paradigm: first they map the problem terrain, then they propose a defensible human-AI planning arrangement for operating within it. But writing the second paper also exposed a further gap that neither paper can fill on its own. Once you have both a pathway and a paradigm, the field still needs a way to compare governance quality across different cases, cities, and planning domains. It needs a way to test whether the mechanisms SPT proposes actually produce their intended results, and under what conditions those mechanisms hold up across different contexts. This comparative and longitudinal challenge is the natural next step, and it points directly to where the research needs to go: developing a shared evaluation infrastructure that can turn AI-enabled planning into a cumulative, testable field of research.
GUIHUA:FURP Editorial Team:
Looking ahead, where does this research program go next? What does it mean for the future of planners and planning education?
Professor Zhong-Ren Peng:
The next phase of this work is no longer really about whether AI can help planning. That question has largely been answered. It is about how AI-enabled planning can stay accountable over time. That is different from initial design or first stage governance. Even a well designed planning system can still drift. Models get retrained, staff turn over, vendor relationships deepen, and power that started as conditional delegation can gradually harden into routine infrastructure. So accountability has to be designed as a lasting institutional condition, not a one time ethical commitment. It should be built into the operating lifecycle of a planning system, not just invoked when the system launches.
For me, two directions matter most. First, treat accountability as infrastructure, not as a vision. The SPT paper shows that procedural compliance can coexist with real inequity for years, with no institutional mechanism able to detect or correct that gap. Preventing this means building governance arrangements in from the start: binding triggers, pathways for contestation, time bound reauthorization, rather than adding these things after the problem has already appeared. The three structural failure modes that worry me most are automation creep, governance decay, and equity backsliding: AI authority quietly expands beyond its original mandate, oversight weakens as the system becomes normalized, and distributional equity erodes slowly under operational pressure. Preventing these foreseeable institutional dynamics requires designing delegation that can be revoked, making sure delegated analytical power remains open to public review, challenge, and withdrawal throughout the system’s operating life.
The second direction is comparative methodology, building the shared evaluation infrastructure this field needs, so research can move from isolated case studies toward cumulative, testable knowledge. Without a common unit of analysis, a shared measurement logic, and portable propositions, a finding in one city cannot inform or challenge a finding in another. This is both a methodological problem and an empirical research problem. Solving it matters a great deal if AI-enabled planning is going to become a cumulative field of research, rather than a growing but mutually incommensurable archive of impressive case studies.
For the planning profession, all of this points to a fundamental shift in the work planners will need to master. The planner of the near future will not mainly be a hands on technician. They will be more like a governance architect, someone who can define goals and guardrails, evaluate and challenge model outputs, design challenge procedures, interpret performance data broken down by equity dimension, and maintain public accountability as the system evolves. This is a demanding set of capabilities, and planning education has not fully caught up. This field needs to develop governance literacy alongside tool literacy. Planners will of course need to understand AI concepts and model evaluation, but they will equally need to understand institutional design, participatory governance, ethics, and how to recognize when a technically capable system is producing substantively unjust outcomes.
The real measure of success for this whole research program is not just whether AI makes planning faster or analytically stronger, though it certainly should do that too. The real measure is this: can cities strengthen their analytical capacity without giving up democratic control over public decisions? In the end, both theory and practice should be judged by that standard.
About GUIHUA
GUIHUA: Frontiers of Urban and Rural Planning is a high-level international academic journal dedicated to cutting-edge research and design in urban and rural planning, documenting the latest developments in urban and rural planning and design in China and around the world for the international academic community.
The journal focuses on ideas, theories, methods, technologies, organization, and governance related to sustainable urban and rural development and planning for a better quality of life. It analyzes and studies urban and rural planning, distills urban and rural knowledge, and explores innovative planning practices, nourishing the discipline through the interaction of academia and practice, and supporting urban and rural development decisions with professional, rational analysis.
Website: https://link.springer.com/journal/44243
Editorial Credits
Editors-in-Chief for this issue: Zhong-Ren Peng, Xiaoxiao Feng
Reviewed by: Xiaoxiao Feng
Promotion: Ruhang Wei