Blue Archive producer warns careless AI use is eroding player trust
A senior developer behind Blue Archive says sloppy use of generative AI is lowering quality and pushing players to question studio integrity. The message is simple: if the work feels cheaper, players assume the intent is too.
Speaking to GameMeca, Yongha Kim, producer at Nexon Games, called out "AI slop" - content rushed through generators without craft or care. In his view, the backlash isn't a misunderstanding of technology. It's a reaction to worse work.
Quality is the issue, not the tech
Kim compared it to deceptive packaging: the bag looks full, but most of it is air. Players notice when the "snacks" - the real creative substance - get reduced. That gap between surface polish and actual value breaks trust fast.
He also pointed to niche, subculture-heavy spaces like gacha, where authorship and voice matter. If the work doesn't feel like it came from a person with intent, the audience tunes out.
"Simulators" can't carry creative intent
Kim described current transformer and diffusion systems as "merely simulators." They remix patterns, not purpose. His question is the one every creative leader should ask: can your intent survive the pipeline, or does it get sanded down into generic output?
How Nexon uses AI (and where they don't)
Kim acknowledged Nexon Games uses AI internally - for speech synthesis, voice recognition, and reducing repetitive tasks. The line is clear: assist production, don't outsource judgment. AI is a tool for throughput, not a replacement for taste.
What this means for creatives
If you lead art, narrative, audio, or UX, the takeaway is practical: protect authorship, keep quality tight, and be transparent about what you automate. You don't need to ban AI. You need to direct it.
Practical guardrails you can implement now
- Define intent first: write a short creative brief before any prompt. If the brief is vague, the output will be too.
- Use AI where repetition drags: cleanup, variations, transcriptions, search, temp VO - not for final vision or style-setting.
- Enforce provenance: keep a changelog for AI-assisted assets and document human review. If you can't explain your process, expect pushback.
- Quality gates over shortcuts: set hard acceptance criteria for AI-assisted work (readability, style match, cohesion, technical fidelity).
- Keep a human at the edges: let people handle first principles (theme, story, composition) and last mile (polish, tone, pacing).
- Respect the subculture: if your audience values a distinct voice, prioritize human-led writing, character art, and worldbuilding.
- Disclose with intent: communicate where AI helps efficiency and where humans lead creative decisions.
A simple workflow that keeps trust intact
- Ideation: human-led themes and constraints → AI for broad exploration → human selects and refines.
- Production: AI for grunt work → human reviews against a style guide → iterate with human edits.
- Final: human polish → peer critique → ship with a brief note on process if relevant.
The bottom line
Players don't reject AI by default. They reject the feeling that the studio values speed over craft. If the work looks good but feels empty, they're gone.
Use AI to clear the busywork. Keep people responsible for taste, authorship, and the final call. That's how you ship faster without burning the trust you need for the long run.
If you're building a responsible workflow across creative roles, explore curated training by job to standardize skills and guardrails: Complete AI Training by Job.
Your membership also unlocks:
