• 5 Posts
  • 493 Comments
Joined 2 years ago
cake
Cake day: July 1st, 2023

help-circle








  • Prioritization probably.

    They are busy keeping up with engineering teams 4x their size for their browser competition… (~500 engineers building Firefox vs ~2000 building chrome)

    Of a million potential features, you can only choose a subset to work on. This one was likely low impact and a low enough use case priority that it got regularly bumped under higher priority work.

    This is just the way… Well… Anything with limited resources works.

    Happy they finally got around for this.










  • I very specifically did not mention LLMs, I even called out that our current technology is not there yet. And llms are current technology.

    The argument in thread was about AGI being impossible or possible. Not necessarily about the articles. Statement of llm-based agis not being possible, which is a pretty obvious and almost unnecessary statement.

    It’s like saying cars with tires and no airfoil surfaces aren’t going to fly. Yeah no shit.

    A fancy text prediction and marginal reasoning engine isn’t going to make AGI. By no means does that make AGI impossible though, since the concept of AGI is not tied to LLMs capabilities