home.social

#dataannotation — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #dataannotation, aggregated by home.social.

fetched live
  1. Computyne launches AI Data Annotation service, bringing 15+ years of BPO expertise to address the critical shortage of quality training data for machine learning development. #AI #DataAnnotation

  2. ‘There’s a lot of desperation’: skilled older workers turn to #AItraining to stay afloat
    5 skilled workers aged 50 and older spoke to Guardian about how, after struggling to find work in their fields, they turned to an emerging and growing category of work: using their expertise to train #artificialintelligence models. Known as #dataannotation, the work involves labeling and evaluating the information used to train #AI models like #OpenAI's #ChatGPT or #Google's #Gemini.
    theguardian.com/technology/ng-

  3. RE: infosec.exchange/@mttaggart/11

    The myth that AI is this magical technology that 'just works' is actively harming tens of thousands of marginalized people. Propagating this idea is harmful. Promoting AI in general is harmful.

    #AI #DataAnnotation #ContentModeration #noAI #slop

  4. Rapidata promises to shrink AI model cycles from months to days by tackling data‑annotation bottlenecks. Their approach blends human feedback, reinforcement learning and robotics to speed up model iteration—backed by ETH Zurich research. Curious how this could reshape open‑source ML pipelines? Read the full story. #AI #DataAnnotation #ModelIteration #Robotics

    🔗 aidailypost.com/news/rapidata-

  5. #Women in rural #India report experiencing #trauma from #dataannotation work, which requires them to review #violentcontent and #porn for global tech companies. The workers describe hours of exposure to #abusivecontent while #training #AI systems, with lasting psychological effects. The investigation highlights the human cost behind AI development and raises questions about labor practices in the tech industry. theguardian.com/global-develop #India #Tech #IndiaTech #Technews

  6. Prompt Engineering vs Fine-Tuning: How AI Models Improve

    If your AI model feels unpredictable, retraining is not always the answer. This article explains prompt engineering vs fine-tuning in plain terms, helping teams get more consistent results, avoid unnecessary costs, and choose the right optimization approach for real-world AI & data annotation services.

    Know More: hitechdigital.com/blog/prompt-

    #PromptEngineering #FineTuning #ArtificialIntelligence #MachineLearning #AIModelData #DataAnnotation

  7. "AI tools have become ubiquitous, entering many facets of everyday life. More often than not, “artificial intelligence” models are presented as fully automated, having dispensed with the need for human intervention. The human workers who train, test, and maintain AI models and act as the first line of defense against model failures are made visible only occasionally. Media coverage sometimes emerges of hundreds of Indian workers1 who remotely ensure the checkout process goes smoothly while creating the illusion of automation at Amazon Go stores and African content moderators2 who make social media platforms safer at great personal cost. But these stories only scratch the surface of the labor that underpins every part of the AI production process.

    Despite being touted as the definitive technological breakthrough of this century, the conditions under which AI models and tools are produced by data workers, in a highly opaque and fissured global supply chain, are still underexplored. Studies of data workers in the Global South have begun to fill gaps in knowledge about the low-paid outsourced labor behind AI, but less is known about U.S. data workers’ conditions.

    In this report, we begin to address this gap through a study of the working conditions of U.S.-based data workers, conducted by AWU-CWA and TechEquity.These workers are essential to the development of tools and models developed by big tech companies, but are employed by complex webs of contractors in the U.S.-based sections of the global AI supply chain. Combining data from a survey of 160 data workers with insights from 15 in-depth interviews, we’ve found that the poor working conditions seen in the Global South are also widespread in data work in the U.S."

    cwa-union.org/ghost-workers-ai

    #DataLabour #DataLabelling #DataAnnotation #BigTech #AI #GenerativeAI #WageSlavery

  8. "Scale AI is basically a data annotation hub that does essential grunt work for the AI industry. To train an AI model, you need quality data. And for that data to mean anything, an AI model needs to know what it's looking at. Annotators manually go in and add that context.

    As is the means du jour in corporate America, Scale AI built its business model on an army of egregiously underpaid gig workers, many of them overseas. The conditions have been described as "digital sweatshops," and many workers have accused Scale AI of wage theft.

    It turns out this was not an environment for fostering high-quality work.

    According to internal documents obtained by Inc, Scale AI's "Bulba Experts" program to train Google's AI systems was supposed to be staffed with authorities across relevant fields. But instead, during a chaotic 11 months between March 2023 and April 2024, its dubious "contributors" inundated the program with "spam," which was described as "writing gibberish, writing incorrect information, GPT-generated thought processes."

    In many cases, the spammers, who were independent contractors who worked through Scale AI-owned platforms like Remotasks and Outlier, still got paid for submitting complete nonsense, according to former Scale contractors, since it became almost impossible to catch them all. And even if they did get caught, some would come back by simply using a VPN.

    "People made so much money," a former contributor told Inc. "They just hired everybody who could breathe.""

    futurism.com/scale-ai-zuckerbe

    #AI #GenerativeAI #Meta #ScaleAI #DataAnnotation #DataLabeling #GigWork

  9. "The production of artificial intelligence (AI) requires human labour, with tasks ranging from well-paid engineering work to often-outsourced data work. This commentary explores the economic and policy implications of improving working conditions for AI data workers, specifically focusing on the impact of clearer task instructions and increased pay for data annotators. It contrasts rule-based and standard-based approaches to task instructions, revealing evidence-based practices for increasing accuracy in annotation and lowering task difficulty for annotators. AI developers have an economic incentive to invest in these areas as better annotation can lead to higher quality AI systems. The findings have broader implications for AI policy beyond the fairness of labour standards in the AI economy. Testing the design of annotation instructions is crucial for the development of annotation standards as a prerequisite for scientific review and effective human oversight of AI systems in protection of ethical values and fundamental rights."

    journals.sagepub.com/doi/10.11

    #AI #GenerativeAI #DataWork #DataLabour #AIPolicy #PoliticalEconomy #DataLabeling #AIEthics #DataAnnotation

  10. 🧬 Can AI fix the chaos in biological sample data?

    🔗 Annotation of biological samples data to standard ontologies with support from large language models. Computational and Structural Biotechnology Journal, DOI: doi.org/10.1016/j.csbj.2025.05

    📚 CSBJ: csbj.org/

    #AIinScience #LLMs #Bioinformatics #DataAnnotation #GPT4 #BiomedicalAI #OpenScience #FAIRData #Ontology #AIinBiology #DataInteroperability

  11. TechXplore: Third-party data annotators often fail to accurately read the emotions of others, study finds. “Machine learning algorithms and large language models (LLMs), such as the model underpinning the functioning of the platform ChatGPT, have proved to be effective in tackling a wide range of tasks. These models are trained on various types of data (e.g., texts, images, videos, and/or […]

    https://rbfirehose.com/2025/05/22/techxplore-third-party-data-annotators-often-fail-to-accurately-read-the-emotions-of-others-study-finds/

  12. Expert #dataAnnotation is tedious and expensive. Can #AI help?

    DANNY's (#Data ANnotation for Non-experts made easY) text and visual feedback improved 28 non-experts' accuracy and AI-reliance.

    #Expertise became necessary as uncertainty rose.

    doi.org/10.1145/3708359.371216

    #medicine