AI Garden Design at the Chelsea Flower Show: What Spacelift Means for Designers and Business

TL;DR: Matt Keightley’s Spacelift—an AI garden design app—produced three full‑size show gardens at Chelsea, triggering a debate that mirrors wider tensions in creative industries: does AI democratise design and improve briefs, or commoditise tacit craftsmanship? Leaders should pilot with clear KPIs, protect premium human services, and set standards for authorship and quality.

What happened at Chelsea—and why it matters for AI garden design

At the Royal Hospital gardens in Chelsea this year, three full‑size show gardens were generated with Spacelift, an AI garden design app created by award‑winning landscape designer Matt Keightley. The gardens included a rural‑inspired build using reclaimed materials, a compact urban balcony solution, and a woodland wellbeing scheme complete with a sauna and cold shower. Presenting AI‑generated designs on such a high‑profile stage forced a fast, public reckoning about the role of automation in a craft that has long been considered hands‑on and place‑sensitive.

Reactions split along predictable lines. The Society of Garden and Landscape Designers (SGLD) criticised the inclusion, arguing that garden design is a collaborative art rooted in human connection. Andrew Duff, chair of the SGLD, said it bluntly:

“Garden design is an art anchored in creativity, collaboration, and human connection; technology can help but cannot replace lived experience and empathy.”

Meanwhile Spacelift’s team and supporters framed the app as a democratising tool that helps homeowners who cannot afford bespoke design and produces clearer briefs for professionals when those clients later hire them.

Quick primer: what the jargon means

• Generative design: software that creates layout options, plant palettes and visual plans from user inputs (photos, measurements, design goals).
• Sensor‑driven monitoring: using soil moisture sensors, sap‑flow gauges, microclimate loggers and similar hardware to produce data that informs planting, irrigation and resilience choices.
• Tacit or embodied knowledge: hands‑on know‑how about soil texture, microclimate quirks, seasonal timing and the social use of a place—skills learned through direct experience rather than explicit rules.

How Spacelift works (practical view)

Spacelift, as showcased, takes a homeowner’s brief and basic site inputs—photos, dimensions, sun/shade data and style preferences—and generates design options, schematic layouts, suggested plant lists and build sequences. The platform also claims to produce practical outputs such as cost ranges and material lists that make a design actionable for DIYers or contractors.

That workflow is important for two reasons. First, it turns vague client wishes into concrete briefs, which can reduce back‑and‑forth and scope creep for professional designers. Second, it creates a low‑cost entry route for customers who otherwise wouldn’t commission a designer at all, expanding the market but also introducing price pressure at the commodity end.

Two distinct AI trends in horticulture

The debate at Chelsea actually highlights two separate technological currents that are often conflated.

• Sensor analytics and AI augmentation: Using sensors and analytics to monitor soil, microclimate and plant health. Designers like Tom Massey use these tools to make better decisions about plant choice, irrigation and long‑term resilience. This is broadly seen as a productivity and quality booster.
• Generative, end‑to‑end design: Platforms that sketch full schemes, select palettes and propose build plans without the same level of on‑site, embodied input. This is more disruptive to the notion of craft and livelihood because it produces a finished intellectual product that can be implemented without a trained designer.

Tom Massey, a Chelsea gold medallist, summarised the distinction: using AI for data and monitoring is different from handing design over to “robot designers.”

Points of contention: craft, access and authorship

There are four core tensions that define the argument:

• Craft vs commodity: Designers worry generative tools will commoditise the years of tacit learning that underpin site‑sensitive design, shrinking margins on routine projects.
• Access vs devaluation: Platforms can bring good design to cost‑conscious homeowners, but wider access could drive down the perceived value of bespoke services.
• Quality and place‑sensitivity: Algorithms can optimise patterns but often miss small‑scale cues—soil quirks, neighbour sightlines, subtle seasonal effects—that determine a garden’s success.
• IP and authorship: If an AI model is trained on existing designers’ work, who owns the output? And when a show like Chelsea displays an AI plan, does it confer cultural legitimacy on machine authorship?

Alexandra Davison of Spacelift frames the platform as improving briefs and access: “The platform is intended to serve homeowners priced out of professional design and to improve the briefing process for the industry.” Keightley himself emphasised practical outcomes:

“The app gives people a practical plan and the confidence to create a garden, bringing technology to a part of the home that has lagged behind.”

Case vignette: a small balcony turned usable

A London homeowner with a narrow 3m balcony used an AI plan as a starting point. The app proposed a layered plant scheme, compact seating and modular planters sized to the exact rail width. The client implemented the layout with a local contractor and reported faster decision‑making and fewer costly mid‑build changes. The designer consulted afterwards to refine hard‑landscape details and plant selection for long‑term resilience.

This mini‑workflow—AI for rapid optioning, human for refinement and hand‑over—illustrates a hybrid model that preserves professional value while improving client access.

Business implications for designers and leaders

For practice leaders and executives in creative services, the Chelsea episode is a case study in choices. The technology can be harnessed to grow markets, improve lead quality and cut repetitive work. Or, if ignored, it can commoditise entry‑level services and accelerate margin pressure.

Key business considerations:

• Market expansion: AI tools widen the pool of customers who can afford a plan. That can generate more projects and referrals if designers position themselves to receive warm leads from DIY or low‑cost clients.
• Operational efficiency: Use AI to automate preliminary concept options, freeing senior staff for higher‑value, site‑specific tasks.
• Pricing strategy: Protect premium services (deep site analysis, bespoke sculptures, historical restoration) with clear value‑based pricing; offer modular services where AI handles entry‑level deliverables.
• Standards and quality control: Agree on verification stamps—e.g., “Human‑Verified Design” or certification tiers—to signal when professional oversight has validated an AI output.
• IP policy: Clarify ownership: does the client own the AI output exclusively, or does the provider retain rights? Ensure contracts cover derivative works and model training concerns.

Key questions executives should ask

• Will AI like Spacelift displace professional garden designers?

  Short term: unlikely to replace bespoke, site‑sensitive professionals. Medium term: lower‑end commissions risk commoditisation unless designers protect premium services and repackage offerings.

• Can AI replicate the tacit, site‑specific knowledge of a designer?

  No—algorithms model patterns and constraints, but they don’t possess hands‑on knowledge of soil texture, microclimate nuances or seasonal rhythms. Human judgment remains essential for those elements.

• Does showing AI work at Chelsea legitimise automated creative outputs?

  It raises public awareness and forces debate. Legitimacy will depend on transparency about methods, perceived quality, and whether professional bodies set standards that preserve trust.

A practical 90‑day pilot checklist for leaders

1. Run a controlled pilot: select 5–10 projects where AI handles concepting and humans do validation. Track time saved, conversion rate and client satisfaction.
2. Define KPIs: average time to concept, lead‑to‑conversion rate, AOV (average order value), margin impact and client NPS.
3. Set IP and contract templates: clarify ownership of AI outputs and permissible reuse.
4. Create a “hybrid offering”: package AI‑generated starter plans with paid professional refinement to protect premium services.
5. Train staff: build quick playbooks for when to accept AI outputs and when to re‑engage senior design judgment.
6. Engage with industry bodies: collaborate on a verification standard or “human‑verified” badge to maintain public trust.

Where to draw the line—and why it matters

Technology has always reshaped creative trades: from drafting tables to CAD, new tools changed workflows but generally enlarged what professionals could offer. The real risk isn’t the existence of AI agents or automation; it’s the pace and framing of adoption. If platforms are integrated thoughtfully—used to improve briefs, expand access, and free human designers for high‑value judgement—everyone benefits. If adoption is unmanaged, the profession risks a race to the bottom on price and a loss of the subtle craft that makes places sing.

The Chelsea debate puts the question plainly: do we let the algorithm hold the trowel, or do we use the algorithm to sharpen the trowel we already hold? Business leaders should answer that deliberately, with pilots, KPIs and clear standards that protect reputation while embracing productive automation.

Alt text suggestion for image: “Spacelift AI garden design exhibit at Chelsea Flower Show — example of AI garden design displayed alongside traditional gardens.”