We don't use AI in any of our design or production processes
A small type design studio’s pledge to avoid AI in its creative and production workflow sparks broader debate over what counts as “proper” tool use in art, design, and software. Some commenters argue that AI is just another productivity aid and that outcomes matter more than process, while others see human-only methods as preserving cultural nuance, craft knowledge, and intentionality that current models cannot replicate. The exchange also highlights growing polarization around AI—between enthusiastic adopters, principled resisters, and a quieter middle who treat it as a limited but useful tool.
Site performance and “hugged to death”
- The site repeatedly went down under HN traffic; some blamed the hosting provider and autoscaling setup.
- Many commenters made jokes that the author should “use AI to fix the site,” while others called this juvenile and argued that a small design studio not hardening for HN-scale traffic is reasonable.
Tone and culture of the discussion
- Several participants lamented a shift on HN toward snark, AI “cult” behavior, and VC/mercenary culture, with less curiosity and good-faith engagement.
- Others pushed back that the article itself had a grandiose or repetitive “AI bad, human good” tone that invited ridicule.
AI as tool vs extremes
- One recurring theme: two “extremes” – people refusing to touch AI on principle vs. people treating it as near-AGI that will solve everything.
- Multiple commenters identified as a quiet middle: using AI selectively, aware it’s powerful but limited, and then moving on with work.
Practical uses and limitations
- Reported productive uses: summarizing and prioritizing email, drafting and smoothing technical texts, coding assistance, home IT and automation projects, and “compressed Stack Overflow.”
- Reported downsides: bad summaries missing crucial details, messy or fragile generated configs/queries, and management pressure to adopt tools that don’t actually save time.
Font design process and reasons to avoid AI
- The author explained that quality typefaces start with analogue sketches; different tools have their own “grain” that shapes outcomes.
- Frequent switching between physical and digital drawing is used to avoid overfitting to one tool’s biases; AI’s “grain” was described as especially unpredictable.
- Manual control and understanding of Bézier curves were framed as essential; delegating this to AI was seen as likely to degrade readability and craft.
Culture, value, and economics
- Some argued that artifacts shape culture, and relying on LLMs risks centering only what appears in training data, excluding many traditions.
- Others responded that every tool and medium has always shaped culture; AI is another step in a long trajectory.
- Concerns surfaced about AI accelerating job loss and wealth concentration, with uncertainty over whether the end state will be more utopian or dystopian.
Reception of the “no AI” stance
- Some saw “we don’t use AI” as a marketing flex that shouldn’t matter if the work is good.
- Others valued explicit commitments to human craft, especially in niche, high-skill fields like type design.