en.Wedoany.com Reported - AI startup Micro AGI has launched a free cleaning service called Shift in New York, aimed at collecting training data to develop humanoid robots. The project deploys two college graduates as cleaners, working five days a week, cleaning approximately five apartments per day, with cameras mounted on the front of their baseball caps to record every action during the cleaning process.
These workers use the collected visual data to train robots to perform household tasks. Shift founder Bercan Kilic told the BBC that the main technical challenge lies in variations in lighting and objects in the real world, and the model needs to learn how hands, cameras, and the environment work together. The project has already collected repair footage from auto mechanics in Turkey and plans to expand into multiple fields such as handymen and maintenance. BBC reporter Archie Mitchell discovered two college graduates in their twenties who came to clean his apartment on the Upper East Side of New York.
This approach of exchanging free services for detailed visual data has raised privacy concerns. Rory Mir, Director of Technology Open Access and Community Engagement at the Electronic Frontier Foundation, warned that although money or services are provided upfront, the shared data could have adverse consequences. Calli Schroeder, Director of the Electronic Privacy Information Center, added that the amount of sensitive information recorded by in-home video footage far exceeds people's estimates.
In-home footage may capture extremely sensitive information such as daily habits and personal documents, and the risk of data breaches or third-party misuse remains. Shift's method of exchanging free services for detailed visual footage reflects the AI industry's hunger for high-quality training data. This real-world interaction data is seen as key to overcoming the limitations of humanoid robots in operating within unstructured environments.
The AI training data market is expected to continue growing rapidly, with companies and startups competing to collect more data to train AI models. Shift's approach of exchanging data for free services could set a precedent for future new business models, but sustainability and privacy regulations are potential obstacles. The collected data, due to its inclusion of real physical interactions with the environment, is considered highly valuable, with its valuation believed to exceed labor costs.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
