home.social

#mercor — Public Fediverse posts

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

  1. All Content from Business Insider | Meta paused its work with AI training startup Mercor after a data breach by Kelsey Vlamis,Charles Rollet

    Meta paused its work with the AI training startup Mercor.Davide Bonaldo/SOPA Images/LightRocket via Getty Images

    Meta has paused its work with Mercor and is investigating after a data breach, a person familiar said.Mercor confirmed to Business Insider that it was recently affected by a security incident.Mercor works with major companies by providing training data for AI models.Meta has paused its work with Mercor and is investigating after a recent data breach at the AI training startup, a source familiar with the matter confirmed to Business Insider.

    Mercor, which was valued at $10 billion in a funding round in October, works with major tech companies like Meta to train AI models with the help of thousands of human contractors and experts. Wired first reported on Friday that Meta had paused all its work with the company.

    Read the original article on Business Insider

    Read more: businessinsider.com/meta-pause

    #meta #mercor #aitraining #litellm

  2. “few other human activities offer such unambiguous feedback. There are no objective tests for whether financial analysis or advertising copy is “good.” Undeterred, AI companies set out to make such tests, collectively paying billions of dollars to professionals of all types to write exacting and comprehensive criteria for a job well done.”

    “… of the more than 30 workers I spoke with occupied a position along a vast and growing data-supply chain. There are people crafting #checklists that define a good #ChatbotResponse, typically called “#rubrics,” and other people grading those rubrics. Others #grade chatbot answers according to those rubrics, and still others take the rubrics and write out what’s often described as a “#GoldenOutput,” or the ideal chatbot answer.”

    “Others are asked to explain every step they took to arrive at this golden output in the voice of a #chatbot thinking to itself, producing what’s called a “#ReasoningTrace” for #AI to follow later when it encounters a similar task out in the real world. Sometimes the labs want only rubrics for prompts their AI can’t already do, which means companies like #Mercor ask workers to produce “#stumpers,” or requests that will make the #AIModel fail.”

    It’s not that AI will totally replace the expertise of #WhiteCollar PhD, Legal and other specialist professions. What will happen is that hard earned expertise will be priced lower. How low? As low as AI technology can make it, think Communist era pay.

    #ZeroHourWork / #work / #economics / #SupplyDemand <theverge.com/cs/features/87738> (paywall) / <archive.md/kAiZL>

  3. “few other human activities offer such unambiguous feedback. There are no objective tests for whether financial analysis or advertising copy is “good.” Undeterred, AI companies set out to make such tests, collectively paying billions of dollars to professionals of all types to write exacting and comprehensive criteria for a job well done.”

    “… of the more than 30 workers I spoke with occupied a position along a vast and growing data-supply chain. There are people crafting #checklists that define a good #ChatbotResponse, typically called “#rubrics,” and other people grading those rubrics. Others #grade chatbot answers according to those rubrics, and still others take the rubrics and write out what’s often described as a “#GoldenOutput,” or the ideal chatbot answer.”

    “Others are asked to explain every step they took to arrive at this golden output in the voice of a #chatbot thinking to itself, producing what’s called a “#ReasoningTrace” for #AI to follow later when it encounters a similar task out in the real world. Sometimes the labs want only rubrics for prompts their AI can’t already do, which means companies like #Mercor ask workers to produce “#stumpers,” or requests that will make the #AIModel fail.”

    It’s not that AI will totally replace the expertise of #WhiteCollar PhD, Legal and other specialist professions. What will happen is that hard earned expertise will be priced lower. How low? As low as AI technology can make it, think Communist era pay.

    #ZeroHourWork / #work / #economics / #SupplyDemand <theverge.com/cs/features/87738> (paywall) / <archive.md/kAiZL>

  4. “few other human activities offer such unambiguous feedback. There are no objective tests for whether financial analysis or advertising copy is “good.” Undeterred, AI companies set out to make such tests, collectively paying billions of dollars to professionals of all types to write exacting and comprehensive criteria for a job well done.”

    “… of the more than 30 workers I spoke with occupied a position along a vast and growing data-supply chain. There are people crafting #checklists that define a good #ChatbotResponse, typically called “#rubrics,” and other people grading those rubrics. Others #grade chatbot answers according to those rubrics, and still others take the rubrics and write out what’s often described as a “#GoldenOutput,” or the ideal chatbot answer.”

    “Others are asked to explain every step they took to arrive at this golden output in the voice of a #chatbot thinking to itself, producing what’s called a “#ReasoningTrace” for #AI to follow later when it encounters a similar task out in the real world. Sometimes the labs want only rubrics for prompts their AI can’t already do, which means companies like #Mercor ask workers to produce “#stumpers,” or requests that will make the #AIModel fail.”

    It’s not that AI will totally replace the expertise of #WhiteCollar PhD, Legal and other specialist professions. What will happen is that hard earned expertise will be priced lower. How low? As low as AI technology can make it, think Communist era pay.

    #ZeroHourWork / #work / #economics / #SupplyDemand <theverge.com/cs/features/87738> (paywall) / <archive.md/kAiZL>

  5. “few other human activities offer such unambiguous feedback. There are no objective tests for whether financial analysis or advertising copy is “good.” Undeterred, AI companies set out to make such tests, collectively paying billions of dollars to professionals of all types to write exacting and comprehensive criteria for a job well done.”

    “… of the more than 30 workers I spoke with occupied a position along a vast and growing data-supply chain. There are people crafting #checklists that define a good #ChatbotResponse, typically called “#rubrics,” and other people grading those rubrics. Others #grade chatbot answers according to those rubrics, and still others take the rubrics and write out what’s often described as a “#GoldenOutput,” or the ideal chatbot answer.”

    “Others are asked to explain every step they took to arrive at this golden output in the voice of a #chatbot thinking to itself, producing what’s called a “#ReasoningTrace” for #AI to follow later when it encounters a similar task out in the real world. Sometimes the labs want only rubrics for prompts their AI can’t already do, which means companies like #Mercor ask workers to produce “#stumpers,” or requests that will make the #AIModel fail.”

    It’s not that AI will totally replace the expertise of #WhiteCollar PhD, Legal and other specialist professions. What will happen is that hard earned expertise will be priced lower. How low? As low as AI technology can make it, think Communist era pay.

    #ZeroHourWork / #work / #economics / #SupplyDemand <theverge.com/cs/features/87738> (paywall) / <archive.md/kAiZL>

  6. “few other human activities offer such unambiguous feedback. There are no objective tests for whether financial analysis or advertising copy is “good.” Undeterred, AI companies set out to make such tests, collectively paying billions of dollars to professionals of all types to write exacting and comprehensive criteria for a job well done.”

    “… of the more than 30 workers I spoke with occupied a position along a vast and growing data-supply chain. There are people crafting #checklists that define a good #ChatbotResponse, typically called “#rubrics,” and other people grading those rubrics. Others #grade chatbot answers according to those rubrics, and still others take the rubrics and write out what’s often described as a “#GoldenOutput,” or the ideal chatbot answer.”

    “Others are asked to explain every step they took to arrive at this golden output in the voice of a #chatbot thinking to itself, producing what’s called a “#ReasoningTrace” for #AI to follow later when it encounters a similar task out in the real world. Sometimes the labs want only rubrics for prompts their AI can’t already do, which means companies like #Mercor ask workers to produce “#stumpers,” or requests that will make the #AIModel fail.”

    It’s not that AI will totally replace the expertise of #WhiteCollar PhD, Legal and other specialist professions. What will happen is that hard earned expertise will be priced lower. How low? As low as AI technology can make it, think Communist era pay.

    #ZeroHourWork / #work / #economics / #SupplyDemand <theverge.com/cs/features/87738> (paywall) / <archive.md/kAiZL>

  7. Laid-off #Scientists and #Lawyers Training #AI to Steal Their #Careers
    #Mercor says 30k work on its platform each week, #ScaleAI claims to have over 70k “M.A.’s, #PhD and college graduates.” Surge AI advertises Supreme Court litigators, McKinsey principals, and platinum recording artists. Experience in #law, #finance, & #coding, areas where AI is making rapid inroads. But also hiring people to produce data for practically any job you can imagine.
    nymag.com/intelligencer/articl
    archive.ph/RVSZM

  8. Wie KI Datenfirmen die Maschine füttern
    Während alle über Rechenzentren sprechen, entsteht im Schatten eine neue KI-Infrastruktur: Datenfirmen, die Arbeitskräfte und Fachwissen in Trainingsdaten verwandeln. Und genau dort fließt derzeit viel Geld.

    Der Markt wächst ra
    apfeltalk.de/magazin/news/wie-
    #Feature #KI #News #Arbeitsmarkt #DataLabeling #HandshakeAI #KI #Mercor #Rechenzentren #RLHF #ScaleAI #Startups #SurgeAI #Trainingsdaten

  9. Sources: AI training startup Mercor eyes $10B+ valuation on $450M run rate

    Mercor, a startup that connects companies like OpenAI and Meta with domain experts needed to train and refine…
    #NewsBeep #News #Entrepreneurship #Business #CA #Canada #datalabeling #FelicisVentures #Mercor #SPV
    newsbeep.com/ca/139372/

  10. Sources: AI training startup Mercor eyes $10B+ valuation on $450M run rate

    Mercor, a startup that connects companies like OpenAI and Meta with domain experts needed to train and refine…
    #NewsBeep #News #Entrepreneurship #AU #Australia #Business #datalabeling #FelicisVentures #Mercor #SPV
    newsbeep.com/au/138450/

  11. Hiring fizzled, execs bailed in months - Meta’s “Superintelligence” dreams are fading as fast as they formed. Maybe cash only goes so far. You’d think #MarkZuckerberg might’ve learned that lesson by now. #AI #GenAI #Meta #MSL #ScaleAI #Mercor #Surge #AlexandrWang

    Cracks are forming in Meta’s p...