Jeremy Merrill
jeremybmerrill.com
Jeremy Merrill
@jeremybmerrill.com
dad, a reporter, data reporter (writing about tech at The Washington Post)

I don't have digital; I don't have diddly squat.

Linguistics, NLP, news, Jews, Atlanta, crypto regs/politics, ads, fraud, etc.
why are you characterizing me as "having a meltdown"?
February 7, 2025 at 8:04 PM
I don't think you've proven your headline.
February 7, 2025 at 2:38 PM
Huh! Fascinating.
January 23, 2025 at 5:29 PM
Reposted by Jeremy Merrill
Fun fact: there used to be a different Gulf of America...in the Sea of Japan archive.org/details/grea...
Great Soviet encyclopedia : Free Download, Borrow, and Streaming : Internet Archive
v. : 29 cm
archive.org
January 23, 2025 at 1:45 AM
Yeah, it's fair question! The view count metrics act right, in that there aren't any sharp discontinuities and many US-focused accounts (i.e. not Elon and Trump) have slower increases overnight and faster increases during the day in the US.
December 20, 2024 at 2:10 PM
I don't think there's any evidence he's just making it up.
December 18, 2024 at 8:47 PM
This was actually one of my research questions for this project: do view count stats over time make sense? And they do! Elon's follow a very consistent pattern. US-centric less-popular figures' view count growth slows overnight and accelerates during awake hours.
December 18, 2024 at 3:27 PM
I was so tempted to buy this when I got the email about it the other day.
December 2, 2024 at 7:26 PM
This one is not actually all that exciting! This is from an examination of TikTok. The x-axis is something like "average proportion of video watched" (I don't totally remember; old chart) and so the big cluster is regular videos and the lil cluster on the left is ads... which on average people skip.
December 2, 2024 at 7:09 PM
Genuinely jazzed to un-forget about this bot, which proposes fixing literally every problem every discussed in (IIRC) 10 years of NYT articles, with crypto/web3/DeFi/Bitcoin.
November 29, 2024 at 7:30 PM