• 6 Posts
  • 2.04K Comments
Joined 2 年前
cake
Cake day: 2023年6月18日

help-circle

  • Bullshit, the dataset is massive and impractical to share, and a dataset may include bias and conditions for use, and the dataset is a completely separate thing from the code. You would always want to use a dataset that fit your needs. From known sources. It’s easy to collect data. Programming a good AI algorithm not so much.
    Saying a model isn’t open source because collected data isn’t included is like saying a music player isn’t open source, because it doesn’t include any music.



  • OpenAI CEO Sam Altman declared a “code red” last week as the upstart faces greater rivalry from Google, threatening its ability to monetize its AI products and meet its ambitious revenue targets.

    Interesting that even Sam Altman is worried now!
    AFAIK there are also problems that Chinese companies have their own tool chain, and are releasing high level truly open source solutions for AI.

    Seems to me a problem for the sky high profits could be that it is hard to make AI lock in, like is popular with much software and cloud services. But with AI you can use whatever tool is best value, and switch to the competition whenever you want.

    It’s nice that it will probably be impossible for 1 company to monopolize AI, like Microsoft did with operating systems for decades.





  • It has to be recording at least for temporary storage to “observe” the surroundings as exemplified.
    Usually the way Google does that is to use a centralized service to treat the data, that means these recordings are also sent to a server.

    This creates immense surveillance potential, and in EU USA and American companies are not considered safe to handle or store such data.

    Even if the recording is deleted seconds after, it is still recording, and “someone” could decide to store it permanently based on the content. For instance based on face recognition.


  • When people do that with smartphones, it’s a conscious decision by the user to record whatever they are recording, and it’s up to the user to decide whether it’s legal.
    Google Glasses record indiscriminately, and will also record illegally, for instance in a situation where there has been an accident, where the people involved have privacy protection. Or if you go into a restaurant, where people also have an expectation of privacy that is protected. Or if you are sitting in your car, which is also protected by privacy.
    So unless the glasses correctly can account for all privacy situations and stop recording, they should be illegal. And since the privacy situation can only be determined AFTER having actually recorded it, I don’t see how they can be legal.
    Except maybe in USA, where laws don’t matter if you are a tech company.


  • You absolutely can. There is no expectation of privacy in any public place except restrooms.

    American standards are third world standards, in civilized countries you cannot just record people, it’s only generally true if you are in a public street, where there absolutely are situations that are protected by privacy such as accidents. Also if you sit in your car or in a restaurant, those are private situations. Glasses that continuously register what you are looking at, are therefore most likely illegal in such situations.








  • I also use KDE, and it is far from minimal, but as I recall my system is only half that with a full system upgrade!
    Some say creativity stuff takes much room, but for instance Blender is only ½ a gig.

    But maybe my system is bigger than I remember, because even at 40 gig it’s near irrelevant compared to the size of an SSD today, and with 1 gigabit internet the upgrades are fast anyway.

    IDK if there’s a way to see the size of my actual Linux install not counting 3rd party media or games?