𝔊𝔴𝔢𝔯𝔫@gwernSep 6My grandmother died of coronavirus+pneumonia today. She had declined ventilators yesterday, and it happened shockingly fast.
We were not close, but ever since February I've been dreading the day when I would finally know someone who died and what Zoom funerals look like. RIP.
143,963
7,590
5.3%
View Tweet activity
𝔊𝔴𝔢𝔯𝔫@gwernSep 6Q. Why did the dying Mr Potato call for a priest?
A. He wished to be chived.
𝔊𝔴𝔢𝔯𝔫@gwernAug 31papers.ssrn.com/sol3/papers.cf…
"NYT...quasi-experiment that varied the 1. quantity (the number of free articles) and 2. exclusivity (the number of available sections) of free content...paywall changes suppressed total content demand by about 9.9%, reducing total advertising revenue."
𝔊𝔴𝔢𝔯𝔫@gwernSep 24But anyway this answers my question: a Hangul-like English font *does* exist, and it's called "Hangulatin"! twitter.com/AnitaJott/stat…fonts.com/font/urw-type-…
Very interesting to look at. As I suspected, parsing English into syllables and compacting into pretty blocks is difficult.
14,251
210
1.5%
View Tweet activity
𝔊𝔴𝔢𝔯𝔫@gwernSep 17Every few years I check en.wikipedia.org/wiki/Berserk_(… to see if it's finished & I can read it yet...
I just noticed it began in 1989 (!). So considering the age distribution for shonen manga, _Berserk_ may now be older than most of its total readership.
12,990
418
3.2%
View Tweet activity
𝔊𝔴𝔢𝔯𝔫@gwernSep 9(This is even more true of GPT & the scaling hypothesis, in case that wasn't obvious.)
12,337
259
2.1%
View Tweet activity
𝔊𝔴𝔢𝔯𝔫@gwernSep 18I think this is a useful trick in many domains: everyone reports nominal amounts even when those are largely irrelevant. For example in investing - how often do we really care about the absolute return or %, compared to the return over alternatives like brain-dead stock indexing?
𝔊𝔴𝔢𝔯𝔫@gwernSep 13reddit.com/r/AIDungeon/co… Who will be the first schizophrenic (or other mental illness, subclinical or not) to go all Terry Davis on a neural language model?
6,904
428
6.2%
View Tweet activity
𝔊𝔴𝔢𝔯𝔫@gwernSep 22"Diffusion/denoising models and latent spaces. You've invented diffusion/denoising models and latent spaces."
5,968
172
2.9%
View Tweet activity
𝔊𝔴𝔢𝔯𝔫@gwernSep 13They have a great one. It's an amazing dataset which allows them to block by city, varying which foreign language wiki is treated to measure nationality effects. That must increase statistical power enormously compared to alternatives like randomizing pairs of cities for totals.
𝔊𝔴𝔢𝔯𝔫@gwernSep 9The wording of the 1k GPU-hours part is confusing: is that 1k per each theorem attempted (so 100k total), or for the entire set of 1k theorems (1 hour each)? twitter.com/gwern/status/1…
5,187
79
1.5%
View Tweet activity
𝔊𝔴𝔢𝔯𝔫@gwernSep 26I don't think there's much to this. It's just Poisson clumping in the equal-odds rule. Same way people see 'hot hands' everywhere.
4,902
162
3.3%
View Tweet activity
𝔊𝔴𝔢𝔯𝔫@gwernSep 1If the people who predicted that this would be wrong & most people would still be working in offices in 2020 now say that they would have been right "but for" X, though, I think I would not hold it against them.
4,624
98
2.1%
View Tweet activity
Get your Tweets in front of more people.
Use Tweet Activity to track how your Tweets are doing.
Engagements
Showing 30 days with daily frequency
Engagement rate
2.6%
Sep 30
1.3% engagement rate
Link clicks
3.6K
Sep 30
20 link clicks
On average, you earned 119 link clicks per day
Retweets without comments
150
Sep 30
0 Retweets without comments
On average, you earned 5 Retweets without comments per day
![](data:image/svg+xml,<svg xmlns="http://www.w3.org/2000/svg" width="128" height="128"><rect fill-opacity="0"/></svg>)
