from Columbia DevFest 2023

Tweeting about Lorentz

I thought I had created something quite interesting, so I felt that if I tweeted it well, it might go viral to some extent.

  • It’s visually interesting.
  • I’m not developing it specifically to go viral, but I’m a little interested in the experience of having a tweet go viral.

So I tried tweeting while coming up with ideas to improve the appearance of the demo video.

Observations:

  • 15 minutes after tweeting:
    • image
      • It’s a speed I’ve never experienced before, 6 retweets in 15 minutes, quite interesting.
        • Except for the first two people, they are not my followers, so the reach seems to be wide.
      • I’m curious to see how the speed will change from here.
  • 30 minutes later:
    • It was retweeted by someone with 30,000 followers, but it doesn’t seem to have a big impact?
    • 11 retweets, 50 likes, 2186 views.
      • Well, it’s about the same speed as before.
  • 60 minutes later:
    • 19 retweets, 70 likes.
      • It’s starting to slow down.
      • I wonder if tweets that go to thousands of retweets will still accelerate.
  • 2 hours later:
    • 39 retweets, 132 likes, 8300 views.
      • Quite linear, very strange.
        • Well, maybe not really.
      • 19 retweets per hour.
      • 75 likes per hour.
  • 3 hours later:
    • 89 retweets, 290 likes, 19000 views.
      • Oh, it’s starting to accelerate more than linearly (blu3mo).
      • Mysterious quoted retweets are starting to come in, 6 of them.
      • For some reason, comments came from the English-speaking community. How did they find it? (blu3mo)
  • 4 hours later:
    • 120 retweets, 420 likes, 29000 views.
  • 15 hours later:
    • 287 retweets, 1117 likes, 116000 views.
      • Well, it’s slowing down a bit.
      • Since it’s late at night in Japan around this time, I guess this is about the limit.

Impressions:

  • Observing it, the growth of retweets and likes is mostly linear (blu3mo).
    • I thought it would be an exponential change based on the exponential function, so it’s strange that it’s linear.
      • If we simplify the model a lot, diffusion should grow exponentially.
        • I thought it would be a finite diffusion due to various factors overlapping.
      • I’m curious if the linearity is just a coincidence or if there’s some explainable phenomenon.
  • The reaction on social media definitely stimulates the desire for approval and distorts people’s evaluation functions/reward systems.
    • I spent quite a bit of time observing the reactions on Twitter, so it’s scary.

/villagepump/2023/02/27

I thought it was high-context (nishio). Understanding “the world looks distorted when moving fast,” understanding that “I created a program in VR space where you can experience that,” understanding that in the forest scene, things are not moving with the naive sense of distance and time, but are actually moving at a speed close to the speed of light and are slowed down so that humans can perceive them, and understanding that when the screen appears to shrink horizontally, you need to understand that you are moving horizontally.

  • That’s true (blu3mo).
    • I also feel that even if people have a perception like “something strange physical phenomenon can be seen in VR,” they might find it interesting.
    • But it seems that I can’t expect diffusion of several thousand to tens of thousands of retweets.