![Nick Kapur Profile](https://pbs.twimg.com/profile_images/1841692292567957504/dB9ZwIeh_x96.jpg)
Nick Kapur
@nick_kapur
Followers
28K
Following
66K
Statuses
26K
Historian of Japan and East Asia. I only tweet extremely interesting things. Also on bluesky - search for my name.
Joined February 2010
RT @shinobu_books: You are not allowed to say the f-word because there are animals without middle fingers. Pictured are some of the animals…
0
8
0
RT @nick_kapur: Of the 24 busiest train stations in the world, the top 23 are all in Japan ❤️🚃🚅🚄🇯🇵
0
3
0
RT @Brian_Riedl: “Why aren’t you cheering Trump & DOGE? I thought you wanted spending and deficit cuts!" Because I’ve been doing this for…
0
2K
0
RT @HistorianZhang: Curious how Chinese chitchat on the US at the moment frequently reference the Cultural Revolution. The extralegal takeo…
0
3
0
RT @modernjapanhist: MJHA is thrilled to announce that the 2025 Modern Japan History Association Book Prize has been awarded to Ryo Morimot…
0
8
0
Is handwriting style a heritable trait? I've seen many cases where children have almost identical handwriting to their parents. Many experts claim this is just because children "subconsciously copy" their parents' handwriting, but this seems to be clearly false. My daughter's handwriting looks uncannily exactly like mine, but in this age of digital texts, she almost never had any chance to see my handwriting, and learned to write entirely at school. It seems genetics must have played a role.
0
0
5
The stock market is freaking out and Nvidia stock droped 17% in a single day yesterday, all because Chinese company DeepSeek found ways to train AI models much more cheaply and using fewer chips. But the notion that training AI more cheaply is bad for Nvidia is ludicrous. AI compute becoming cheaper will only drive more demand for AI compute, from a wider variety of actors, which will only increase demand for Nvida chips, in a classic case of the Jevons paradox. And DeepSeek's AI models were all trained on...Nvidia chips.
3
0
11