home.social

#ainews — Public Fediverse posts

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

  1. #OpenAI adjusted several #evaluationbenchmarks for its #GPT6 #Astra model after the initial announcement, with some metrics showing Astra performing better and others worse. The changes, which included altering hallucination rates and maths scores, sparked concerns about #benchmaxxing and the accuracy of #benchmark tests. fortune.com/2026/09/04/openai- #tech #news #ainews

  2. People are sharing their #darkestsecrets with #AIchatbots, unaware that these conversations can be accessed by tech companies, law enforcement, and even employers. Despite the perception of #privacy, chatbot transcripts have been used in #court cases, and there’s a risk of #dataleaks from AI companies. Users are often unaware that their conversations are used for #trainingAI models and can be #subpoenaed. futurism.com/artificial-intell #tech #news #ainews

  3. #Texas #Republicans are #opposing the rapid expansion of #AI #datacentre|s, citing concerns about rising costs and strained resources. This shift in attitude, driven by local opposition, poses a challenge for both big tech and Trump’s push for AI dominance. techradar.com/pro/texas-republ #tech #news #ainews

  4. #Texas #Republicans are #opposing the rapid expansion of #AI #datacentre|s, citing concerns about rising costs and strained resources. This shift in attitude, driven by local opposition, poses a challenge for both big tech and Trump’s push for AI dominance. techradar.com/pro/texas-republ #tech #news #ainews

  5. #Texas #Republicans are #opposing the rapid expansion of #AI #datacentre|s, citing concerns about rising costs and strained resources. This shift in attitude, driven by local opposition, poses a challenge for both big tech and Trump’s push for AI dominance. techradar.com/pro/texas-republ #tech #news #ainews

  6. #Texas #Republicans are #opposing the rapid expansion of #AI #datacentre|s, citing concerns about rising costs and strained resources. This shift in attitude, driven by local opposition, poses a challenge for both big tech and Trump’s push for AI dominance. techradar.com/pro/texas-republ #tech #news #ainews

  7. #Texas #Republicans are #opposing the rapid expansion of #AI #datacentre|s, citing concerns about rising costs and strained resources. This shift in attitude, driven by local opposition, poses a challenge for both big tech and Trump’s push for AI dominance. techradar.com/pro/texas-republ #tech #news #ainews

  8. 📰 DeepSeek and Qwen just dropped input prices, while Optuna v5 lands. Two big model cost cuts and a new hyperparameter tool in one day—see the tracking data.

    olud.ai/news/2026-09-07.html

    #AI #OpenSource #AINews

  9. 📰 DeepSeek and Qwen just dropped input prices, while Optuna v5 lands. Two big model cost cuts and a new hyperparameter tool in one day—see the tracking data.

    olud.ai/news/2026-09-07.html

    #AI #OpenSource #AINews

  10. 📰 DeepSeek and Qwen just dropped input prices, while Optuna v5 lands. Two big model cost cuts and a new hyperparameter tool in one day—see the tracking data.

    olud.ai/news/2026-09-07.html

    #AI #OpenSource #AINews

  11. 📰 DeepSeek and Qwen just dropped input prices, while Optuna v5 lands. Two big model cost cuts and a new hyperparameter tool in one day—see the tracking data.

    olud.ai/news/2026-09-07.html

    #AI #OpenSource #AINews

  12. 📰 DeepSeek and Qwen just dropped input prices, while Optuna v5 lands. Two big model cost cuts and a new hyperparameter tool in one day—see the tracking data.

    olud.ai/news/2026-09-07.html

    #AI #OpenSource #AINews

  13. #AIcompanies are rapidly releasing #newmodels and updates, creating complexity for users trying to keep up. This #modelfatigue is driven by companies vying for market share in a rapidly growing #AImarket projected to reach $2.59 trillion this year. While the pace of innovation is exciting, concerns are rising about the potential dangers of increasingly capable AI models and the lack of clear regulation. cnbc.com/2026/09/06/meta-googl #tech #news #ainews

  14. #AIcompanies are rapidly releasing #newmodels and updates, creating complexity for users trying to keep up. This #modelfatigue is driven by companies vying for market share in a rapidly growing #AImarket projected to reach $2.59 trillion this year. While the pace of innovation is exciting, concerns are rising about the potential dangers of increasingly capable AI models and the lack of clear regulation. cnbc.com/2026/09/06/meta-googl #tech #news #ainews

  15. #AIcompanies are rapidly releasing #newmodels and updates, creating complexity for users trying to keep up. This #modelfatigue is driven by companies vying for market share in a rapidly growing #AImarket projected to reach $2.59 trillion this year. While the pace of innovation is exciting, concerns are rising about the potential dangers of increasingly capable AI models and the lack of clear regulation. cnbc.com/2026/09/06/meta-googl #tech #news #ainews

  16. #AIcompanies are rapidly releasing #newmodels and updates, creating complexity for users trying to keep up. This #modelfatigue is driven by companies vying for market share in a rapidly growing #AImarket projected to reach $2.59 trillion this year. While the pace of innovation is exciting, concerns are rising about the potential dangers of increasingly capable AI models and the lack of clear regulation. cnbc.com/2026/09/06/meta-googl #tech #news #ainews

  17. #AIcompanies are rapidly releasing #newmodels and updates, creating complexity for users trying to keep up. This #modelfatigue is driven by companies vying for market share in a rapidly growing #AImarket projected to reach $2.59 trillion this year. While the pace of innovation is exciting, concerns are rising about the potential dangers of increasingly capable AI models and the lack of clear regulation. cnbc.com/2026/09/06/meta-googl #tech #news #ainews

  18. #SwissRe’s #JérômeHaegeli states that the »#physicaleconomy is back« due to a surge in #datacentre #investment, driving a global #capitalexpenditure #supercycle. This boom, spanning data centres, energy systems, and other infrastructure, presents a significant insurance opportunity, potentially generating $200 billion in premiums between 2026 and 2030. reinsurancene.ws/physical-econ #tech #news #ainews

  19. #SwissRe’s #JérômeHaegeli states that the »#physicaleconomy is back« due to a surge in #datacentre #investment, driving a global #capitalexpenditure #supercycle. This boom, spanning data centres, energy systems, and other infrastructure, presents a significant insurance opportunity, potentially generating $200 billion in premiums between 2026 and 2030. reinsurancene.ws/physical-econ #tech #news #ainews

  20. #SwissRe’s #JérômeHaegeli states that the »#physicaleconomy is back« due to a surge in #datacentre #investment, driving a global #capitalexpenditure #supercycle. This boom, spanning data centres, energy systems, and other infrastructure, presents a significant insurance opportunity, potentially generating $200 billion in premiums between 2026 and 2030. reinsurancene.ws/physical-econ #tech #news #ainews

  21. #SwissRe’s #JérômeHaegeli states that the »#physicaleconomy is back« due to a surge in #datacentre #investment, driving a global #capitalexpenditure #supercycle. This boom, spanning data centres, energy systems, and other infrastructure, presents a significant insurance opportunity, potentially generating $200 billion in premiums between 2026 and 2030. reinsurancene.ws/physical-econ #tech #news #ainews

  22. #SwissRe’s #JérômeHaegeli states that the »#physicaleconomy is back« due to a surge in #datacentre #investment, driving a global #capitalexpenditure #supercycle. This boom, spanning data centres, energy systems, and other infrastructure, presents a significant insurance opportunity, potentially generating $200 billion in premiums between 2026 and 2030. reinsurancene.ws/physical-econ #tech #news #ainews

  23. #Spammers are using a technique called #ASCIIsmuggling, which involves embedding invisible #Unicodetags into #spammessages to evade filters. This technique, originally used to hide malicious prompts in AI attacks, allows spammers to #camouflage #triggerwords and evade detection by machine learning and natural language processing models. Microsoft provided guidance on how developers can better account for it in #spamfilters. arstechnica.com/security/2026/ #tech #news #ainews

  24. #Spammers are using a technique called #ASCIIsmuggling, which involves embedding invisible #Unicodetags into #spammessages to evade filters. This technique, originally used to hide malicious prompts in AI attacks, allows spammers to #camouflage #triggerwords and evade detection by machine learning and natural language processing models. Microsoft provided guidance on how developers can better account for it in #spamfilters. arstechnica.com/security/2026/ #tech #news #ainews

  25. #Spammers are using a technique called #ASCIIsmuggling, which involves embedding invisible #Unicodetags into #spammessages to evade filters. This technique, originally used to hide malicious prompts in AI attacks, allows spammers to #camouflage #triggerwords and evade detection by machine learning and natural language processing models. Microsoft provided guidance on how developers can better account for it in #spamfilters. arstechnica.com/security/2026/ #tech #news #ainews

  26. #Spammers are using a technique called #ASCIIsmuggling, which involves embedding invisible #Unicodetags into #spammessages to evade filters. This technique, originally used to hide malicious prompts in AI attacks, allows spammers to #camouflage #triggerwords and evade detection by machine learning and natural language processing models. Microsoft provided guidance on how developers can better account for it in #spamfilters. arstechnica.com/security/2026/ #tech #news #ainews

  27. #Spammers are using a technique called #ASCIIsmuggling, which involves embedding invisible #Unicodetags into #spammessages to evade filters. This technique, originally used to hide malicious prompts in AI attacks, allows spammers to #camouflage #triggerwords and evade detection by machine learning and natural language processing models. Microsoft provided guidance on how developers can better account for it in #spamfilters. arstechnica.com/security/2026/ #tech #news #ainews

  28. #Discord is implementing #ageverification for accessing age-restricted content, stage channels, and changing DM settings. The #verificationprocess, which is already enforced in some regions, will be globally implemented by the second half of 2026. Users can choose from four verification methods, each with varying #dataprivacy implications. gadgetreview.com/discord-is-br #tech #news #ainews

  29. #Discord is implementing #ageverification for accessing age-restricted content, stage channels, and changing DM settings. The #verificationprocess, which is already enforced in some regions, will be globally implemented by the second half of 2026. Users can choose from four verification methods, each with varying #dataprivacy implications. gadgetreview.com/discord-is-br #tech #news #ainews

  30. #Discord is implementing #ageverification for accessing age-restricted content, stage channels, and changing DM settings. The #verificationprocess, which is already enforced in some regions, will be globally implemented by the second half of 2026. Users can choose from four verification methods, each with varying #dataprivacy implications. gadgetreview.com/discord-is-br #tech #news #ainews

  31. #Discord is implementing #ageverification for accessing age-restricted content, stage channels, and changing DM settings. The #verificationprocess, which is already enforced in some regions, will be globally implemented by the second half of 2026. Users can choose from four verification methods, each with varying #dataprivacy implications. gadgetreview.com/discord-is-br #tech #news #ainews

  32. #Discord is implementing #ageverification for accessing age-restricted content, stage channels, and changing DM settings. The #verificationprocess, which is already enforced in some regions, will be globally implemented by the second half of 2026. Users can choose from four verification methods, each with varying #dataprivacy implications. gadgetreview.com/discord-is-br #tech #news #ainews

  33. #AItokens, units of computing used by AI models, are becoming increasingly common in #China. Businesses are offering them as #incentives, including credit card rewards, mobile data plans, freebies at restaurants and cafes, and even as a form of #currency on e-commerce platforms. This trend is driven by the popularity of AI and the competitive pricing of Chinese AI models, making tokens more accessible to consumers. restofworld.org/2026/china-ai- #tech #news #ainews

  34. #AItokens, units of computing used by AI models, are becoming increasingly common in #China. Businesses are offering them as #incentives, including credit card rewards, mobile data plans, freebies at restaurants and cafes, and even as a form of #currency on e-commerce platforms. This trend is driven by the popularity of AI and the competitive pricing of Chinese AI models, making tokens more accessible to consumers. restofworld.org/2026/china-ai- #tech #news #ainews

  35. #AItokens, units of computing used by AI models, are becoming increasingly common in #China. Businesses are offering them as #incentives, including credit card rewards, mobile data plans, freebies at restaurants and cafes, and even as a form of #currency on e-commerce platforms. This trend is driven by the popularity of AI and the competitive pricing of Chinese AI models, making tokens more accessible to consumers. restofworld.org/2026/china-ai- #tech #news #ainews

  36. #AItokens, units of computing used by AI models, are becoming increasingly common in #China. Businesses are offering them as #incentives, including credit card rewards, mobile data plans, freebies at restaurants and cafes, and even as a form of #currency on e-commerce platforms. This trend is driven by the popularity of AI and the competitive pricing of Chinese AI models, making tokens more accessible to consumers. restofworld.org/2026/china-ai- #tech #news #ainews

  37. #AItokens, units of computing used by AI models, are becoming increasingly common in #China. Businesses are offering them as #incentives, including credit card rewards, mobile data plans, freebies at restaurants and cafes, and even as a form of #currency on e-commerce platforms. This trend is driven by the popularity of AI and the competitive pricing of Chinese AI models, making tokens more accessible to consumers. restofworld.org/2026/china-ai- #tech #news #ainews

  38. 📰 Today's AI brief: LiteLLM signs images, DeepSeek cuts prices, Qwen3.5 jumps

    New models, notable releases, price moves and emerging projects - measured from our own tracking, published daily.
    olud.ai/news/2026-09-06.html

    #AI #OpenSource #AINews

  39. 📰 Today's AI brief: LiteLLM signs images, DeepSeek cuts prices, Qwen3.5 jumps

    New models, notable releases, price moves and emerging projects - measured from our own tracking, published daily.
    olud.ai/news/2026-09-06.html

    #AI #OpenSource #AINews

  40. 📰 Today's AI brief: LiteLLM signs images, DeepSeek cuts prices, Qwen3.5 jumps

    New models, notable releases, price moves and emerging projects - measured from our own tracking, published daily.
    olud.ai/news/2026-09-06.html

    #AI #OpenSource #AINews

  41. 📰 Today's AI brief: LiteLLM signs images, DeepSeek cuts prices, Qwen3.5 jumps

    New models, notable releases, price moves and emerging projects - measured from our own tracking, published daily.
    olud.ai/news/2026-09-06.html

    #AI #OpenSource #AINews

  42. 📰 Today's AI brief: LiteLLM signs images, DeepSeek cuts prices, Qwen3.5 jumps

    New models, notable releases, price moves and emerging projects - measured from our own tracking, published daily.
    olud.ai/news/2026-09-06.html

    #AI #OpenSource #AINews

  43. The #hacker group #Rhysida leaked almost six terabytes of data from #Berlin’s state administration on the dark web, including #sensitiveinformation about CBRN threats, national defence plans, and personal data of state civil servants. The Berlin Senate refused to pay the demanded ransom of 30 Bitcoin. euronews.com/next/2026/09/05/b #tech #news #ainews

  44. The #hacker group #Rhysida leaked almost six terabytes of data from #Berlin’s state administration on the dark web, including #sensitiveinformation about CBRN threats, national defence plans, and personal data of state civil servants. The Berlin Senate refused to pay the demanded ransom of 30 Bitcoin. euronews.com/next/2026/09/05/b #tech #news #ainews

  45. The #hacker group #Rhysida leaked almost six terabytes of data from #Berlin’s state administration on the dark web, including #sensitiveinformation about CBRN threats, national defence plans, and personal data of state civil servants. The Berlin Senate refused to pay the demanded ransom of 30 Bitcoin. euronews.com/next/2026/09/05/b #tech #news #ainews

  46. The #hacker group #Rhysida leaked almost six terabytes of data from #Berlin’s state administration on the dark web, including #sensitiveinformation about CBRN threats, national defence plans, and personal data of state civil servants. The Berlin Senate refused to pay the demanded ransom of 30 Bitcoin. euronews.com/next/2026/09/05/b #tech #news #ainews