AI tools and analogue craft are reshaping UX/UI design. Here's how Southeast Asian brands can turn the tension between them into measurable commercial value.
There’s a productive contradiction sitting at the centre of design right now. On one side: AI-assisted workflows promising to compress hours of work into minutes. On the other: designers like James Kuwamura deliberately slowing down — reaching for Risograph printers, layering analogue textures, making work that you can almost feel through the screen. Both impulses are rational. The mistake is treating them as opposites rather than a spectrum you can position your brand along — deliberately, and with commercial intent.
The Terminal as Creative Tool Has a Revenue Problem
Pablo Stanley’s exploration of terminal-based design — using command-line environments as a creative canvas — surfaces something important beyond the novelty. When designers move computation closer to their creative process, they gain precise control but introduce a new cost: legibility. Stakeholders, brand managers, and clients in markets like Thailand or Indonesia, where design decisions often travel through multiple layers of approval, cannot interrogate a bash script the way they can a Figma frame.
The practical implication: highly technical design toolchains compress individual output but can create organisational bottlenecks. For teams shipping product UI across Shopee storefronts or LINE mini-apps, the constraint isn’t creative horsepower — it’s how fast a design decision can get cross-functional sign-off. Building human-readable documentation and design tokens into any advanced workflow isn’t optional overhead; it’s what makes the speed gain actually land in production.
Analogue Craft as a Differentiation Signal — With Measurable Impact
Kuwamura’s Risograph project — homages to the three national teams tied to his layered heritage — does something commercially interesting without trying to. By leaning into tactile, imperfect, physically-grounded aesthetics, it produces work with a provenance that AI generation currently cannot replicate convincingly. The layering of personal history and heritage gives the output a specificity that audiences read as authenticity.
For brands operating in Southeast Asia, this is worth translating directly. Regional consumers — particularly across Vietnam, the Philippines, and Malaysia — respond measurably to design that signals local cultural fluency, not generic pan-Asian visual language. A fintech brand in Kuala Lumpur that commissions original batik-influenced motion graphics instead of licensing stock illustration isn’t just making an aesthetic choice. It’s making a trust signal. The conversion rate difference between a landing page that reads as genuinely local versus one that reads as templated is not small — A/B tests in e-commerce contexts routinely show 15–25% lift from culturally-resonant imagery over generic alternatives.
The implementation question is resource allocation. Commissioning original analogue-influenced work costs more upfront but amortises well when it anchors a design system that scales across digital touchpoints. The failure mode to avoid: treating it as a one-off brand campaign asset rather than a reusable system component.
AI in the Dashboard: The Token Cost Nobody Is Budgeting For
Speckyboy’s analysis of WordPress 7.0’s Connectors API puts a sharp edge on the AI-in-design conversation. The core argument: not every workflow task justifies AI, and the economic logic depends heavily on token costs that most teams aren’t tracking. Content summarisation, alt-text generation, and basic SEO metadata are genuinely time-positive use cases. Automated design decisions — layout generation, brand voice enforcement — are where the ROI gets murky fast.
For digital teams managing multi-market campaigns across Southeast Asia, the token cost question compounds. Running AI-assisted content workflows across six language variants (Bahasa Indonesia, Thai, Tagalog, Vietnamese, Malay, English) means token consumption scales with linguistic complexity, not just content volume. A regional brand publisher running 200 content pieces a month at six localisations each is looking at a materially different AI infrastructure cost than their single-market counterpart. The strategic call is to identify the three to five workflow tasks where AI compression is genuinely high-value — metadata, image alt text, first-draft brief generation — and treat the rest as human-owned, because that’s also where the brand differentiation lives.
The broader principle: AI tools are excellent at eliminating the mechanical tax on creative work. They are poor substitutes for the judgment calls that make design commercially distinctive.
Designing the Tension Into Your Brand System
The most sophisticated position available to design teams right now isn’t choosing AI efficiency or analogue craft — it’s architecting a brand system that uses both deliberately. Automated design tokens and component libraries handle the high-frequency, low-differentiation work: button states, form fields, notification UI. Human authorship — illustration style, editorial photography direction, motion language — is reserved for the touchpoints that carry brand meaning.
Shopee’s seller ecosystem does this implicitly. Platform UI is ruthlessly templated for conversion efficiency. But the campaigns that move product — particularly during 11.11 or Hari Raya — rely on bespoke creative that no template generates. The distinction is intentional: efficiency infrastructure enabling human creative investment where it actually compounds.
The implementation path for most teams starts with an audit: map every design output against two axes — frequency of production and contribution to brand distinctiveness. High frequency, low distinctiveness: automate. Low frequency, high distinctiveness: protect the human hours.
Key Takeaways
- Analogue-influenced design signals authentic cultural fluency in Southeast Asian markets — and that signal has measurable conversion value when built into a scalable design system rather than treated as campaign decoration.
- AI workflow tools in CMS and dashboard environments carry token costs that scale with multilingual operations; budget for them explicitly or they erode the efficiency gains they promise.
- The commercially smart move is not AI versus human craft, but a deliberate architecture that automates the mechanical and protects human authorship at the brand-differentiating touchpoints.
The underlying question for design leaders heading into the second half of 2026 is not whether to adopt AI tools — that decision is largely made. It’s whether your brand system is explicit about where human authorship lives, and whether that intentionality is legible to the people consuming your design. In markets where trust is built through cultural recognition as much as product quality, the answer to that question is a revenue number, not just a philosophy.
At grzzly, we help brands across Southeast Asia architect design systems that are operationally efficient and commercially differentiated — not one at the expense of the other. If you’re trying to draw that line inside your own stack, we’d enjoy the conversation. Let’s talk
Sources
Written by
Inkblot GrizzlyCrafting dashboards that tell the truth, and monetisation frameworks that make that truth commercially useful. Turns abstract data assets into revenue-generating products for publishers and brands alike.