The New Price of a Clean Home: Your Privacy
Tech startups are offering free house cleaning in exchange for video data to train robots.
🕒 生成時間: (台北時間)
Summary · 摘要
A German startup called MicroAGI is offering free home cleaning services in New York City. In exchange, the company records the cleaners to collect data for training household robots. This trend highlights the growing need for physical-world data in the artificial intelligence industry. While the company promises to protect privacy through blurring, experts remain concerned about data security. Other companies are also exploring similar ways to gather footage of human chores.
一家名為 MicroAGI 的新創公司正在紐約市提供免費的居家清潔服務。作為交換,該公司會記錄清潔人員的工作過程,以收集訓練居家機器人的數據。此趨勢凸顯了人工智慧產業對現實世界數據的需求日益增長。儘管該公司承諾透過模糊處理來保護隱私,但專家仍對數據安全感到擔憂。其他公司也在探索類似的方式來收集人類家務活動的影像。
Imagine a stranger coming into your home to clean your kitchen, fold your laundry, and mop your floors for free. It sounds like a dream for anyone who dislikes chores. However, there is a catch. A new tech startup called MicroAGI is offering this service in New York City, but with one major condition: the cleaners must wear cameras to record everything they do. This footage is then used to teach artificial intelligence (AI) how to perform physical tasks in the real world.
According to Ars Technica, the company launched its "Shift" app on May 28. The service is part of a larger mission to improve "embodied AI," which refers to machines that can move and interact with physical objects. By recording cleaners as they scrub dishes or dust tables, the company hopes to create a database that will eventually allow robots to do these jobs for us. The Verge reports that this is a difficult challenge because, unlike internet text or images, the physical world is complex. Robots struggle to understand things that humans do naturally, such as picking up an apple or folding clothes.
This need for "real-world data" has created a new kind of business. Because it is hard to find enough video of people doing housework, companies are getting creative. The Verge notes that while AI models like chatbots were easily trained using data taken from the internet, physical robots require high-quality video of human movement. This has led to a "bottleneck"—a situation where progress is slowed down because there is not enough information available. To solve this, some startups are paying people to wear cameras, while others are creating "data farms" where workers are paid to repeat simple tasks like folding towels over and over so that sensors can capture every movement.
Privacy is the biggest concern for many people. The Shift app website claims that it uses advanced technology to automatically blur faces, screens, and personal documents before any data is uploaded to its servers. The company says this process is "irreversible," meaning the information cannot be turned back into a clear image. However, Ars Technica points out that there is no clear way for customers to ask for their videos to be deleted later. It is also not certain if these blurring techniques are enough to keep a person’s home from being identified in the long run.
MicroAGI is not the only company interested in this data. In India, a home services platform called Pronto has also been using client homes to record chores like cooking and laundry. This practice has caused a wave of negative reactions in the industry. According to The Verge, some rival companies have stated that they would never record inside a customer's home, arguing that it crosses a line. Meanwhile, other businesses, such as Human Archive, are trying to scale up this process by partnering with workers who wear camera-equipped hats to record their daily activities from their own point of view.
What comes next for these robots? Despite the excitement in the tech world, true automation—where a robot can fully manage a household without help—is still a long way off. For now, companies are focused on gathering as much data as possible. While the idea of a robot doing the dishes is appealing, the current method of collecting that data forces homeowners to decide how much of their private life they are willing to share. As these startups expand to other cities like London, the debate over whether the convenience of a free cleaning is worth the loss of privacy will likely continue to grow.
選擇題練習 · Quiz
共 4 題
- 細節 Detail
1.According to the article, what is the primary technical difficulty in training robots to perform household chores?
- 推論 Inference
2.What can be inferred about the companies collecting data in people's homes?
- 單字情境 Vocabulary
3.In the third paragraph, what does the word 'bottleneck' refer to in the context of AI development?
- 主旨 Main Idea
4.What is the central message of the article regarding the use of AI in household chores?
易誤解詞彙 · Words to watch
這些字字面意思和文中用法不同,或是不常見的詞性/片語。
- catch noun
- A hidden problem or disadvantage in a situation that seems good.
- 隱藏的缺點、陷阱或條件。
- 💡 常見作動詞(捕捉),這裡作名詞,指隱藏的條件或問題。文中:However, there is a catch.
- bottleneck noun
- A situation that causes progress to slow down or stop.
- 瓶頸;阻礙進展的情況。
- 💡 原意指瓶子的頸部,這裡引申為進展受阻的關鍵點。文中:This has led to a "bottleneck"—a situation where progress is slowed down because there is not enough information available.
- crosses a line idiom
- To behave in a way that is considered unacceptable or immoral.
- 越界;做出無法接受的行為。
- 💡 字面意思是跨越界線,這裡指行為超過了道德或社會容忍的限度。文中:According to The Verge, some rival companies have stated that they would never record inside a customer's home, arguing that it crosses a line.
- scale up phrasal verb
- To increase the size, amount, or importance of something.
- 擴大規模;增加產量。
- 💡 常見於商業情境,指將某項業務或流程擴大。文中:Meanwhile, other businesses, such as Human Archive, are trying to scale up this process by partnering with workers who wear camera-equipped hats to record their daily activities from their own point of view.
原始來源 · Sources
本文內容由 AI 從以下來源綜合改寫。事實請以原始來源為準。
gemini/gemini-3.1-flash-lite-preview