• 0 Posts
  • 1.5K Comments
Joined 1 year ago
cake
Cake day: July 2nd, 2023

help-circle






  • The push to require paying to read content that is shared openly on the web is designed to drive a feasibility wedge between large and small operations.

    It means that a small handful of organizations will actually have the funds to be able to buy enough training data, and that all other smaller AI ventures will be illegal.

    This is designed to concentrate AI power into a few hands.

    Think critically about the narratives being fed to you. AIs must be allowed to read the web, because if they are not then we will have a unipolar AI ecosystem and the future of humanity will be extremely dark.

    We need a multipolar AI ecosystem and the only way to do that is to have lots of small players making their own AI.

    If we have a multipolar AI ecosystem, then AI will be forced to play nice because of the effects of parity, and therefore AI will develop along a prosocial, negotiating, respectful path.

    Unipolar AI will be tyrannical, cruel, and evil. Unipolar AI will be the result of making it illegal to train AI on web content.

    Please see past the propagandistic narrative. Today it is OpenAI that is fighting for this right. Tomorrow it will be smaller players. That is a good thing. That is what we want.









  • My potential argument against it starts with asking where the credentials are stored for authenticating this identity.

    Currently the home instance stores the hashed password and performs authentication.

    In a way, the identity “belongs to” the place that does authentication, which now happens to be the instance.

    If identity is decoupled from an instance, that means authentication decouples from an instance.

    If the identity “belongs to” the fediverse as a whole, then that means the fediverse as a whole has an authentication mechanism.

    Unless we can come up with a distributed authentication mechanism, that means the fediverse as a whole has some authentication service, as in one, which means centralized.

    This therefore breaks decentralization, unless the authentication is somehow handled in a distributed way. Maybe consensus or something on a hashed password? But if those hashed passwords are stored in a distributed manner, then you’d need a super long password to prevent rainbow table attacks on the passwords, given the hashed values would essentially be public information.

    Maybe public keys are stored in a blockchain? I don’t know this is beyond me in the details.

    But to summarize the problem at a data model level, an identity belongs to an instance, because the instance can authenticate them. If the identity now belongs to the whole fediverse, then the whole fediverse needs to be able to authenticate them, which if not handled correctly could lead to centralized authentication, centralized banning, censorship, reddit, etc.


  • Yes. If you know the material’s molecular composition, you can tune a series of lasers to break those bonds with minimal energy input.

    You don’t even need the explicit makeup of the material. You can just do a quick calibrating scan consisting of a rainbow sweep of every frequency, while watching to see the scattering pattern of the photons.

    Assuming that your laser tuning system makes different tones when it’s producing different wavelengths of light, your weapon’s appearance while firing would be like:

    • It makes a sound like VWEEEP and plays a shifting rainbow pattern across the object to be disrupted
    • Once calibration is complete, it makes a loud BWAAAAAH sound as the object turns to vapor/powder

    The calibration is only like one second. These properties conveniently make it awesome to include in a cinematic depiction such as a sci fi movie.