Physics and Free Software

  • 20 Posts
  • 530 Comments
Joined 1 year ago
cake
Cake day: June 5th, 2023

help-circle



  • People talk about being there “for” someone. Here being there “with” him is more important. Something as mundane as letting him take a nap on your couch while you fold laundry. The more regular, boring or routine the better. Best if it’s something you can do without talking. That’s harder since you can’t go see each other, but maybe just leaving a video call on or suggesting it to someone else would help.



  • There are worse things in our daily lives. Like the giant pothole you hit on the way to the store. Your neighbors dog shitting in your yard. The stack of TPS reports your asshole boss left on your desk. These have a direct negative impact on your life. Regardless of their agenda, politicians will always have less impact on us than our daily does of regular human bullshit.

    Pessimistic as that may be, there’s an upshot. As many times as that dog shits on your grass, you might watch your favorite baseball team win. Or for every TPS report, you have a real sincere laugh at the water cooler. Think about the scope. Keep it in perspective.

    Note: This is meant for the people who leaned hard into the move to Canada meme last time. Obviously, there are people who’s lives will be irrevocably be changed. Check your privilege and fight for those who lack it. People of conviction don’t hide.




  • Training a model takes more power than what? Generating a single poem? Using it to generate an entire 4th grade class’s essays? To answer all questions in Hawaii for 6th months? What is the scale? The break even point for training is far far less than total usage.

    Have you ever used one locally? Depending on your hardware it’s anywhere between glacially to a morgue’s AC slow. To the average person on the average computer it is nearly unusable, relative to the instant gratification of the web interface.

    That gives you a sense of the resources required to do the task at all, but it doesn’t scale linearly. 2 computers aren’t twice as fast as one. It’s logarithmic. With diminishly returns. In the end, this means one 100 word response uses the equivalent of 3 bottles of water.

    How many queries are made per hour? How does that scale over time with increased usage of the same model? More than training a model. A lot more.