Figma Make, Lovable, and the Race to the Gesamtapp
Figma’s next move At its Config keynote, Figma surprised no one—and yet startled everyone—by unveiling Figma Make, an AI-assisted app-builder that turns design artifacts (or even a text prompt) into working, deploy-ready code. The beta launches alongside three sibling products—Sites, Buzz, and Draw—positioning Figma as a full-stack creation suite rather than “just” a UI collaboratory.
Why now? Lovable and “vibe-coding.” The immediate catalyst appears to be Lovable.dev, the buzzy “idea-to-app-in-seconds” platform where developers trust the AI’s vibes more than they read the code. Lovable markets itself as a fix for vibe-coding’s reliability woes—security, testing, and deploy pipelines baked in. Figma already owned the spec-to-mockup hand-off; Lovable threatened to own the mockup-to-product leap. Make is Figma’s counter-punch.
The Great Convergence: From Point Tools to Gesamtapps
Wagner’s Gesamtkunstwerk—the “total work of art”—is a useful metaphor for what’s happening across creative software. Every category leader is metastasizing into an all-in-one environment:
| Origin Tool | “First Purpose” | Convergent Add-Ons |
|---|---|---|
| DaVinci Resolve | Color grading | Two distinct NLEs (Cut & Edit), Fusion VFX, Fairlight audio post, cloud review hub |
| Blender | 3-D DCC | 2-D Grease Pencil illustration, video editing, game-engine experiments, USD pipeline |
| Unity | Real-time game engine | Muse generative AI, asset store, cinematic tools |
| Unreal | Real-time game engine | MetaHuman Creator, Fab marketplace, UEFN, broadcast graphics |
| Adobe Creative Cloud | Image editing (Photoshop) | Firefly text-to-image/video, Bulk Create, Express social kit |
| Canva | Templated graphics | Magic Studio (text→design/video), Affinity & Leonardo.ai acquisitions |
| AI chatbots | Text convo | Canvas for code & docs, image generation, scheduled tasks—yet still trapped in a single text pane |
Add Framer, Webflow, Notion, Obsidian, VS Code + Copilot, and the picture is clear: silos are eroding, toolchains are collapsing inward, and every vendor wants to own the entire creative/developer pipeline.
Where Are the Choke Points?
- Asset gravity – Once your design files, color grades, or 3-D rigs live in one namespace, switching feels prohibitively expensive.
- Mental-model lock-in – Each app drags its original worldview with it. Resolve still looks like a colorist’s timeline. ChatGPT still behaves like a chat bubble even when you’re editing multiline code.
- Interface extensibility – Can the surface area stretch without tearing? Figma’s canvas was built for components and variant logic; that scaffolding adapts to states, interactions, and ultimately executable code.
- Ecosystem effects – Plugins, template markets, and learning communities snowball advantage (see Unreal’s new Fab marketplace).
Whichever platform turns these pinch-points into seamless on-ramps will define the default Gesamtapp.
Why Graphic-Design Thinking Might Win
Graphic design sits at a crossroads of content, structure, and interaction. A designer:
- Composes spatial hierarchies (layout grid → UI constraints → DOM / components).
- Balances text, imagery, and motion—the same primitives you need in apps, games, and marketing sites.
- Iterates visually at the speed required for modern product cycles.
That cognitive toolkit maps cleanly onto app-building once the rendering and logic layers are abstracted by AI. By contrast:
- Chat-first UIs struggle once tasks exceed linear conversation.
- Game engines excel at 3-D worlds but feel heavy for CRUD dashboards.
- Video editors are timeline-centric: great for sequences, awkward for stateful interactivity.
Figma’s canvas already teaches users component abstraction, variant logic, and tokenized design systems—the very muscles we flex when architecting software. Add Make’s code synthesis and the leap from design intention to runnable app becomes almost legible.
Open Questions
- Multiplicity or monopoly? Will we juggle 30 competing Gesamtapps, or does one platform accrue decisive network effects?
- Which origin worldview scales best? Is it easier to bolt logic onto visuals (Figma), visuals onto logic (Lovable), or everything onto conversation (OpenAI)?
- What happens to the specialist tool? Does the Photoshop-grade colorist still lift Resolve’s Color page, or does one-click Firefly bulk edits quietly suffice for 80 % of work?
- Regulatory & IP pressure? If governments demand audit trails for AI-generated code/art, whose pipeline is auditable end-to-end?
A Bet on Figma (for Now)
If we define a Gesamtapp as the place where:
- Ideation, design, production, and iteration live in a single file format;
- AI agents scaffold boilerplate but leave room for craft;
- Collaboration primitives (comment, branch, version, hand-off) are native,
then Figma is unusually well-positioned. It already owns the live-multiplayer doc for interface work—precisely the bridge discipline between storytelling and engineering. Make simply extends the bridge to the other shore.
But software history tells us that bridges collapse, markets swerve, and “first-purpose” inertia cuts both ways. The next 18 months will reveal whether Figma’s canvas can stretch into a full-stack stage, or whether we’ll be lured elsewhere by a different vibe.
Which choke-point do you think matters most—and which worldview would you trust to build your next everything?
- created by o3