Skip to content

Massive internet outage reported: Google services, Cloudflare, Character.AI among dozens of services impacted

Technology
140 95 102
  • Large Language Model Performance Doubles Every 7 Months

    Technology technology
    53
    1
    96 Stimmen
    53 Beiträge
    13 Aufrufe
    V
    in yes/no type questions, 50% success rate is the absolute worst one can do. Any worse and you're just giving an inverted correct answer more than half the time
  • China is rushing to develop its AI-powered censorship system

    Technology technology
    2
    1
    39 Stimmen
    2 Beiträge
    6 Aufrufe
    why0y@lemmy.mlW
    This concept is the enemy of the a centuries old idealistic societal pillar of the West: Liberté, Libertas... this has blessed so many of us in the West, and I beg that it doesn't leave. Something beautiful and as sacred as the freedom from forced labor and the freedom to choose your trade, is the concept of the free and unbounded innocence of voices asking their leaders and each other these questions, to determine amongst ourselves what is fair and not, for our own betterment and the beauty of free enterprise. It's not so much that the Chinese state is an awful power to behold (it is and fuck Poohhead)... but this same politic is on the rise in the West and it leads to war. It always leads to war. And now the most automated form of state and corporate propaganda the world has ever seen is in the hands of a ruthless ruling class that can, has, and will steal bread from children's hands, and literally take the medicine from the sick to pad their pockets. Such is the twisted fate of society and likely always will be. We need to fight and not with prayers; this moment is God forsaking us to behold how the spirit breaks and what the people want to fight for as ruthlessly as the others do to steal our bread.
  • 32 Stimmen
    6 Beiträge
    30 Aufrufe
    G
    Yes. I can't imagine that they will go after individuals. Businesses can't be so cavalier. But if creators don't pay the extra cost to make their models compliant with EU law, then they can't be used in the EU anyway. So it probably doesn't matter much. The Llama models with vision have the no-EU clause. It's because Meta wasn't allowed to train on European's data because of GDPR. The pure LLMs are fine. They might even be compliant, but we'll have to see what the courts think.
  • WhatsApp rolls out AI-generated summaries for private messages

    Technology technology
    26
    1
    96 Stimmen
    26 Beiträge
    64 Aufrufe
    W
    So I think, but I'm not sure, this is for group chats. Group chats are only encrypted to/from the server because the server broadcasts the message to each recipient. As the messages are unencrypted on the server, they can feed them to LLMs. This is different to Signal. On Signal it's your phone encrypting each copy of the message before sending to each recipient individually.
  • Ads on YouTube

    Technology technology
    47
    30 Stimmen
    47 Beiträge
    13 Aufrufe
    K
    this is like a soulless manager or some ai bot trying to figure why the human brain hates terrible interruptions
  • 55 Stimmen
    4 Beiträge
    18 Aufrufe
    M
    Tragedy of the commons? Everyone wants to use it, no one wants to put forward the resources to maintain it.
  • 33 Stimmen
    4 Beiträge
    8 Aufrufe
    A
    Phew okay /s
  • Apple Eyes Move to AI Search, Ending Era Defined by Google

    Technology technology
    2
    10 Stimmen
    2 Beiträge
    14 Aufrufe
    ohshit604@sh.itjust.worksO
    It’s infuriating that Safari/Apple only allows me to choose from five different search engines. I self-host my own SearXNG instance and have to use a third-party extension to redirect my queries.