The German startup MicroAGI has initiated a program in New York City offering complimentary residential cleaning services in exchange for the collection of first-person audiovisual data.

德國新創公司 MicroAGI 在紐約市啟動了一項計劃，提供免費的居家清潔服務，以換取第一視角的視聽數據。

The operational framework of the 'Shift' application involves the deployment of professional cleaners equipped with head-mounted recording devices. This methodology is designed to capture 'egocentric' data—specifically the navigation of physical space and the execution of domestic tasks—to facilitate the development of embodied AI and household robotics. The company asserts that the intrinsic value of this training data justifies the subsidization of the cleaning services.

''Shift'' 應用程式的運作框架涉及部署配備頭戴式錄影設備的專業清潔人員。此方法旨在捕捉''自我中心''數據——特別是物理空間的導航和居家任務的執行——以促進具身 AI 和家用機器人的開發。該公司聲稱，這些訓練數據的內在價值足以支持清潔服務的補貼。

Regarding data governance, MicroAGI claims the utilization of machine learning models to execute irreversible anonymization, such as the blurring of faces and personally identifiable information, prior to cloud upload. However, the absence of a defined mechanism for data deletion from training sets presents a potential point of contention. Furthermore, the company's terms of service include clauses that indemnify the platform against liability for property damage or personal injury.

在數據治理方面，MicroAGI 聲稱在上傳至雲端前，利用機器學習模型執行不可逆的匿名化處理，例如模糊化面孔和個人識別資訊。然而，缺乏從訓練集中刪除數據的明確機制，這可能成為爭議點。此外，該公司的服務條款包含免除平台對財產損失或人身傷害責任的條款。

This initiative exists within a broader industry trend toward the monetization of real-world physical data. Similar strategies have been observed with entities such as Encord, Micro1, and Pronto, the latter of which encountered market friction following the disclosure of its data collection practices in India. MicroAGI's strategy further encompasses a global recruitment effort, employing tens of thousands of 'operators' across 15 countries to record routine activities for financial compensation. The organization intends to expand these operations to additional urban centers, including London, Munich, and Zurich, while diversifying the scope of data collection to include plumbing and culinary tasks.

此舉屬於將現實世界物理數據貨幣化的更廣泛行業趨勢。在 Encord、Micro1 和 Pronto 等實體中也觀察到類似策略，其中後者在印度披露其數據採集做法後遇到了市場摩擦。MicroAGI 的策略還涵蓋全球招聘，在 15 個國家雇用了數萬名''操作員''，透過金錢補償記錄日常活動。該組織打算將這些業務擴展到其他城市中心，包括倫敦、慕尼黑和蘇黎世，同時將數據採集範圍擴大至水電維修和烹飪任務。

MicroAGI is currently leveraging free service offerings to acquire the high-fidelity physical data necessary to overcome existing bottlenecks in robotic AI development.

MicroAGI 目前利用免費服務獲取高保真物理數據，以克服機器人 AI 開發中現有的瓶頸。