• 0 Posts
  • 74 Comments
Joined 1 year ago
cake
Cake day: July 6th, 2023

help-circle





  • That’s immensely expensive though, and not guaranteed to work because much of that stuff is still research stage. You’re right that paring down the models to make them leaner and more specialized is the primary direction that current research is pursuing, but it’s far from certain at this point how to do it, how well it will work, and how small you can get them before they start to fall apart. Not something game studios are likely to gamble their budgets on, at least not yet.

    We’re nowhere near the “just hire a guy to trim it down instead of hiring writers” stage, and it’s unclear yet whether or not that’s where we’ll end up. We could pull off “just hire a guy to fine-tune an existing foundation model,” but that doesn’t make them smaller.












  • “Legitimate interest” is a concept from GDPR, which is the EU’s major legislation around digital rights. “Legitimate interest” is an extremely soft concept that basically says “a company must have some reason that is not obviously bullshit to process your data.” That includes advertising. “We need to know your age, gender and city so we can decide what ads to show you” is considered “legitimate interest.” See? Soft.

    What it doesn’t cover is “We just think it’s desirable to accumulate as much data as we can about you for no particular reason and maybe we can just sell it one day, idk.” That would not be considered “legitimate interest.” Similarly, asking for e.g. a user’s phone number but having no particular explanation for why you need it would not be “legitimate interest.”

    So when you see “legitimate interest” cookies as a category, you can interpret that as “cookies that have some purpose-- including advertising-- as opposed to literally no purpose other than superstitious data hoarding.” Block them.


  • Those kinds of patterns are already emerging! That “mulling the result through a loop” step is called “reflection,” and it does a great job of catching mistakes and hallucinations. Nothing is on the scale of doing the whole problem-solving and implementation from business requirements to deployed product-- probably never will be, IMO-- but this “making the LLM a component in a broader system with diverse tools” is definitely something that we’re currently figuring out patterns for.