[–] 0 points 2 years ago

No it wouldn’t have.

Licensing isn’t 1 sided. You have to have both parties AGREE to the license.

In this case it’s just text on each message.

The reason OpenAI was able to scrape the internet largely consequence free is because they never agreed to the ToS of the websites, thus they are not bound by those terms.

I agree that it isn’t right or morally correct, but legally it’s okay.

Additionally, by signing up to Lemmy, they grant certain rights and privileges to their instance and others due to agreeing to the license.

In fact, lemmy could make it easier for scrapers, as lemmy will happily deliver content to your instance, simplifying scraping.

  • source
  • parent
  • context
  • [–] 3 points 2 years ago (1 child)

    Look for military surplus stores. Military surplus stores are hit or miss on availability and quality, since they buy in bulk containers. Check over the piece of clothing thoroughly for damage.

    You can also buy some of coats brand new from manufacturers like London Fog. They make an English officers coat that is lovely

  • source
  • parent
  • context
  • [–] 3 points 2 years ago

    The 3rd party investigation found no evidence.

    That doesn’t mean it’s true or not true.

    In these types of situations, especially years afterwards, there is a huge lack of evidence, people forget, systems change and data is lost, etc.

    The 3rd party investigation is done to reduce liability, not determine the truth.

  • source
  • parent
  • context
  • [–] 1 point 2 years ago

    It doesn’t make sense to make battle droids hyper intelligent. They are supposed to be mass produced, why would you make them expensive?

    We see that they can make hyper intelligent droids, TCW has assassin droids and commander droids.

    They are clearly more expensive, since some are uniquely adorned with decorations.

    Astromechs would have to be on the more intelligent side, since they handle flying, mechanic work, navigation, etc.

    The average B1 droids doesn’t need much more than chatgpt.

  • source
  • parent
  • context
  • [–] 4 points 2 years ago

    It’s because LLMs don’t work with letters. They work with tokens that are converted to vectors.

    They literally don’t see the word “strawberry” in order to count the letters.

    Splitting the letter probably separates them into individual tokens

  • source
  • parent
  • context
  • view more: next ›