# NewEyes 用摄像头取代提示词，实现主动式多模态 AI

- 来源：Rohan Paul (@rohanpaul_ai)
- 发布时间：2026-08-05 00:34
- AIHOT 分数：35
- AIHOT 链接：https://aihot.virxact.com/items/cmsevqd8d18iuro2eb9djjwua
- 原文链接：https://x.com/rohanpaul_ai/status/2084679429037633898

## AI 摘要

NewEyes 尝试从多模态 AI 中移除提示词，让摄像头本身成为触发机制：视觉感知识别潜在需求，持久记忆补充上下文，系统自行判断是否采取行动。该应用由 CollovLabs 推出，是首批由端侧多模态智能驱动的消费级 AI 应用之一，可主动实现所见即购、穿搭建议、房间设计、餐食计数及植物缺水提醒等功能。

## 正文

NewEyes is trying to remove the prompt from multimodal AI.

Instead of opening an app， describing a task， and attaching an image， the camera itself becomes the trigger.

Visual perception identifies a possible need， persistent memory adds context， and the system decides whether anything should happen next.

Which shows up as shopping what you see， styling what you wear， redesigning your room， counting your meals， and flagging a thirsty plant before you ask.

This is a meaningful interface shift， because prompts currently tell the model both what matters and what the user permits it to do.

A proactive system has to infer the first while remaining conservative about the second.

Once the balance is achieved， the camera moves from a feature to the everyday layer， where personal AI is actually there.

### 引用推文

> Neweyes AI (by Collov Labs)：What if your camera could... 🛍️ Shop what you see. 👗 Style what you wear. 🛋️ Design the room around you. 🍽️ Count the calories on your plate. 🪴 Remind you ...