All articles tagged with #data labeling
Meta is drafting about 7,000 employees into AI-focused groups—Applied AI, Agent Transformation Accelerator, and Agent Data & Optimization—as part of CEO Mark Zuckerberg's push to accelerate the company's AI efforts, with participation non-negotiable and roles likely involving data labeling, amid ongoing layoffs and workforce reshuffles.
A WIRED tester tries DoorDash’s Tasks app, which pays people to film themselves performing everyday activities to generate training data for AI and robotics. Tasks include laundry, cooking eggs, and park walks, with upfront rates around $15/hour but actual earnings prove meager (under $10 after a few tasks) due to strict filming rules and time caps. The app also restricts certain locations and raises concerns about privacy in public spaces, the sustainability of low-wage gig work for AI development, and the broader implications for workers in the AI data‑labeling ecosystem.
Swedish reports say offshore Meta workers in Kenya are labeling intimate and disturbing videos captured by Ray‑Ban smart glasses to train AI, including footage from bathrooms and clips showing personal data. Working under Sama, they describe being required to review such material, raising privacy and consent concerns as Meta pushes live AI features and potential facial recognition, amid related moderator‑work lawsuits.
Utkarsh Amitabh, a 34-year-old entrepreneur and academic, earns $200/hour training AI models for micro1, leveraging his expertise in business, tech, and philosophy. He values the role for its intellectual challenge and flexibility, viewing AI's impact on jobs with cautious optimism, believing in a future of human-AI collaboration. The job highlights the growing importance of domain experts in AI development and the evolving landscape of work in the AI era.
A shadow market for illicit sale of verified AI training accounts has emerged on social media platforms like Facebook, WhatsApp, and Telegram, driven by the high demand for data labeling work in AI development. Despite company bans and safeguards, scammers and opportunists are selling fake or real accounts, risking scams and data leaks for buyers and sellers. Companies like Scale AI and Surge AI are actively fighting fraud, but the underground industry continues to evolve, highlighting the challenges in securing AI training data.
Scale AI is laying off 14% of its staff, mainly in data-labeling, after rapid growth and a recent $14.3 billion deal with Meta. The company is shifting focus to enterprise and government sales following the loss of major clients, reflecting a strategic pivot in response to market changes.
Uber is expanding its AI data labeling services through Uber AI Solutions, aiming to compete in the market following Scale AI's deal with Meta. The company is offering ready-to-use datasets, licensing its internal platforms, and developing automation tools to streamline project setup. With a presence in over 30 countries and a growing network of clickworkers, Uber seeks to leverage its size and experience to establish a significant position in the AI data labeling industry, amidst increased competition and industry shifts.
OpenAI is ending its collaboration with Scale AI, a data-labeling startup, following Meta's significant investment and hiring of Scale's founder. This move reflects OpenAI's shift towards more specialized data providers to support advanced AI models, amidst concerns about Meta's stake revealing competitors' AI efforts. The decision marks a change from earlier plans to continue working with Scale, as OpenAI seeks better data expertise for developing human-like reasoning AI.
Meta has finalized a $14.3 billion investment to acquire a 49% stake in Scale AI, valuing the startup at $29 billion, with CEO Alexandr Wang joining Meta to lead its AI strategy, signaling a strategic move to enhance its AI capabilities and gain insights into competitors' data priorities.
Scale AI's $15 billion deal to sell a 49% stake to Meta and Wang's departure to lead a new Meta AI lab raises concerns about trust and data sharing with major clients like OpenAI, potentially threatening Scale's business. The deal could also shift the competitive landscape in AI data services, with rivals positioning themselves as neutral alternatives. The move signifies Meta's intensified focus on AI development, but also introduces risks for Scale's existing operations and client relationships.
Meta is in talks to invest over $10 billion in Scale AI, a key player in AI data labeling and model training, which has grown significantly and influenced US AI policy. Scale's evolution from image labeling to advanced language models and its deepening ties with the US government and Meta highlight its central role in the AI industry, driven by the need for high-quality data and expert input.
Meta is in talks to invest over $10 billion in Scale AI, making it one of the largest external AI investments, to support data labeling services for AI training, with Scale AI expected to generate $2 billion in revenue this year.
Tesla employees privately shared invasive videos and images recorded by customers’ car cameras, including embarrassing situations, crashes, and road-rage incidents, according to interviews with nine former employees. Some recordings were made when cars were parked and turned off, potentially revealing where a Tesla owner lived. The sharing of sensitive videos illustrates one of the less-noted features of artificial intelligence systems: they often require armies of human beings to help train machines to learn automated tasks such as driving. Tesla increasingly has been automating the process, but it continues to employ hundreds of data labelers in Buffalo, New York.