AlterU AlterU Part 3 · Vision Part 3 · Vision
大纲Outline
能力 · recognize (Vision)

让 AI 看玩家拍的照。

  • 输入：玩家拍照 → upload → recognize
  • 输出：labels / attributes / parts / caption / confidence
  • 结构化结果再喂给 LLM 出文字 / gen-image 出风格化图
  • 身份钩 ③（环境钩）的引擎——AI 看你的东西 = 心理拥有感最强
  • 已落地：Trash or Treasure / Fit Check / Field Guide / Mugshot Booth / Pet Filter

Vision 把游戏从"AI 给你看东西"翻转成"你给 AI 看东西"—— 这个翻转改变玩家心理位置，传播性天然强。

Trash or Treasure Fit Check
Capability · recognize (vision)

Let AI look at what the player shot.

  • Input: player shoots photo → upload → recognize
  • Output: labels / attributes / parts / caption / confidence
  • Feed the structured result into LLM for words and gen-image for stylized art
  • Engine for identity hook ③ (environment) — AI looking at your stuff = strongest sense of ownership
  • Shipped: Trash or Treasure / Fit Check / Field Guide / Mugshot Booth / Pet Filter

Vision flips "AI shows you things" into "you show AI things." That flip changes the player's psychological position — and tends to come with built-in virality.

Trash or Treasure Fit Check

备忘 · 按 S 关闭Notes · S to close

Vision 是 AlterU 未来主攻方向之一。它独有的是玩家主动给 AI 拍东西这一动作——主动 = 强承诺 = 强情感投入。

所以 Vision 类游戏几乎天然有传播性，玩家会自己截屏发朋友圈。Trash or Treasure / Fit Check 都是这个范式的产物。

Vision is one of AlterU's main bets for the next stretch. The thing nothing else has is the player aiming a camera at the AI — active commitment, strong investment, share-friendly output.

15 / 41