Favorite Rejects 2021-09-17

Noah Smith 🐇 @Noahpinion

Whoa

Hong Kong Free Press HKFP @hkfp

Covid-19: Hong Kong lawmaker Regina Ip gets BioNTech jab after no antibodies detected from Sinovac vaccines https://t.co/snPwnvHswz https://t.co/uB0WroYHdw

Matthew Yglesias @mattyglesias

This is why they should have explicitly made one mRNA vaccine the Republican one and the other the Democratic one.

Curtis Houck @CurtisHouck

Joy Reid again defends her belief that it was NECESSARY to be anti-vaxxer when Trump was in office: "I was hesitant. When Donald Trump was out there controlling the CDC & controlling the FDA & manipulating them & making them put out falsehoods, anybody rational was hesitant." https://t.co/dBFT8lTvWw

Matthew Yglesias @mattyglesias

It is good that he said this but also kind of disturbing that it needs to be set. A key to rebuilding institutional confidence is for the experts themselves to stay in their lanes.

Apoorva Mandavilli @apoorva_nyc

And he is also reminding the advisers to stay focused on science rather than on operational issues of rollout or questions of global equity. Interesting.

Matthew Yglesias @mattyglesias

Tales of Norse exploration of North America made it to 14th century Italy. https://t.co/VM9cIHyPba

Ross Douthat @DouthatNYT

This is cool (via @tylercowen): https://t.co/EfvSQK8qWD

Matthew Yglesias @mattyglesias

There are valid reasons to present a stock/flow ratio but it doesn’t make sense to say things like “Apple’s market cap is bigger than the GDP of Bulgaria” or whatever. https://t.co/t1iB32fCOV

Alec Stapp @AlecStapp

Two things I've changed my mind on in recent years: Fahrenheit is in fact better than Celsius You can (and sometimes should) compare stocks and flows

Ryan Radia @RyanRadia

@mattyglesias Looks like Apple's market cap is roughly 12 times the national net wealth of Bulgaria ($2.42 trillion vs $204 billion).

Andy Slavitt 🇺🇸💉 @ASlavitt

BREAKING: 100% yes vote. Vaccine boosters approved on EUA for those 65 and over and at high risk.

Timothy B. Lee @binarybits

If you’d told an AI researcher in 2001 that AI systems in 2021 would be about as smart as the average republican voter, they would have been pretty impressed. https://t.co/1kKrlJmShu

Owain Evans @OwainEvans_UK

Paper: New benchmark testing if models like GPT3 are truthful (= avoid generating false answers). We find that models fail and they imitate human misconceptions. Larger models (with more params) do worse! PDF: https://t.co/3zo3PNKrR5 with S.Lin (Oxford) + J.Hilton (OpenAI) https://t.co/QfwokYJ7Hq