Sam Atis's Avatar

Sam Atis

@samatis.bsky.social

hello blue sky

372 Followers  |  1,060 Following  |  19 Posts  |  Joined: 06.11.2024  |  2.2249

Latest posts by samatis.bsky.social on Bluesky

The common narrative that AI is underwhelming relative to hype is missing the forest for the trees on the pace of progress. Timelines are shortening.

20.05.2025 19:56 β€” πŸ‘ 30    πŸ” 4    πŸ’¬ 1    πŸ“Œ 1

I think you would expect a result like this even in the absence of any causal effect.

04.12.2024 11:59 β€” πŸ‘ 2    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

Seems like there are plenty of attacks that just wouldn't work (and so aren't attempted) because people would care that they're false. Nobody (/very few people) is going to bother saying 'AOC has literally murdered hundreds of people and most Democrats are serial killers'.

04.12.2024 10:30 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Post image

🚨 NEW PAPER: When low-income Americans get $1,000/month for 3 years, what happens to their political views & behavior?

The OpenResearch Unconditional income Study reveals surprising findings about the effects of income on politics... 🧡

02.12.2024 19:00 β€” πŸ‘ 207    πŸ” 70    πŸ’¬ 11    πŸ“Œ 11
Post image Post image Post image Post image

Do people become more likely to vote when they get more cash?

And what about their political beliefs? Do they change with more money?

This new working paper uses a unique RCT and finds no effects of cash on political participation or engagement.

Political preferences also largely do not change.

02.12.2024 14:58 β€” πŸ‘ 22    πŸ” 5    πŸ’¬ 2    πŸ“Œ 1

How do you weight the wellbeing of the people migrating relative to the wellbeing of people from the UK? Are you saying this is negative overall or just for the UK?

28.11.2024 17:32 β€” πŸ‘ 3    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

Why?

28.11.2024 17:03 β€” πŸ‘ 2    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

πŸ‘€

28.11.2024 14:19 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

Bremen is one of Germany's few cities providing data to track spatial inequality in voter turnout over a long time span. In the 1980s, turnout varied little by neighborhood class composition. More than 30 years later, historic working-class neighborhoods now exhibit markedly lower turnout rates. πŸ‘‡

27.11.2024 07:32 β€” πŸ‘ 43    πŸ” 12    πŸ’¬ 2    πŸ“Œ 0

According to me go.bsky.app/D34oYEo

25.11.2024 22:29 β€” πŸ‘ 21    πŸ” 4    πŸ’¬ 7    πŸ“Œ 1

8 or 9

25.11.2024 10:14 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Post image

A thread about being wrong:

5 years ago, we wrote a paper about how how newly enfranchised 16-year-olds vote in Austria. But we were wrong.

This year, @elisabethgraf.bsky.social, @schnizzl.bsky.social, Sylvia Kritzinger and I are setting the record straight: authors.elsevier.com/c/1juT5xRaZk...

21.11.2024 18:00 β€” πŸ‘ 640    πŸ” 156    πŸ’¬ 26    πŸ“Œ 54
Preview
β€˜I got XL bullies banned – now I am setting my sights on jailing career criminals for good’ Dr Lawrence Newport, who exposed death surge by bully dogs, campaigns to push Government for tougher approach on crime

Dr Lawrence Newport is Britain’s most effective campaigner.

After getting XL Bullies banned, he’s turned his sights on the criminal justice system and the need to focus resources on β€œprolific offenders”.

www.telegraph.co.uk/news/2024/11...

23.11.2024 13:41 β€” πŸ‘ 54    πŸ” 14    πŸ’¬ 4    πŸ“Œ 0
Post image

Kieran Healy is a prince amongst #sociologist, and we should all work up the courage to follow his example #AcademicWriting #AcademicChatter

22.11.2024 07:50 β€” πŸ‘ 96    πŸ” 17    πŸ’¬ 4    πŸ“Œ 5
Post image

Yet it doesn't take a genius to realise class does not mean what it used to. Traditional labels like "working" or "middle" class labels tell us increasingly little about someone's economic situation.

Most people in "professional" jobs say they are working class...

2/4

21.11.2024 20:20 β€” πŸ‘ 10    πŸ” 2    πŸ’¬ 1    πŸ“Œ 1
Preview
Examining the replicability of online experiments selected by a decision market - Nature Human Behaviour This study finds that decision markets can be a useful tool for selecting studies for replication. For a sample of 26 online experiments published in PNAS selected by a decision market, the authors fi...

New in NHB: We systematically replicated 26 of 41 PNAS social science studies that used MTurk. The approach provides a proof-of-concept for using decision markets to select findings to replicate.

Replication effect sizes were 45% of original effect sizes.

www.nature.com/articles/s41...

19.11.2024 11:59 β€” πŸ‘ 100    πŸ” 55    πŸ’¬ 2    πŸ“Œ 1

Yeah, this one seems weird though. It doesn’t actually matter what I call it, so why insist on Claude? It’s not like ChatGPT is being dishonest here.

18.11.2024 23:19 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

Ok!

18.11.2024 23:06 β€” πŸ‘ 2    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

Say more?

18.11.2024 23:05 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

True :(

18.11.2024 23:03 β€” πŸ‘ 5    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0
Post image Post image

Claude has a bad vibe

18.11.2024 22:51 β€” πŸ‘ 13    πŸ” 0    πŸ’¬ 3    πŸ“Œ 2

I made a forecasting and prediction markets starter pack.

Who am I missing? (Probably a lot of people, I don't know who is here) go.bsky.app/TGQ4Rxt

18.11.2024 17:51 β€” πŸ‘ 17    πŸ” 2    πŸ’¬ 1    πŸ“Œ 0

If we don’t understand sentience, how do you know AI is not going to be sentient?

17.11.2024 18:05 β€” πŸ‘ 2    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

What’s your most dunkable take? I’ve never seen you post something patently ridiculous - now’s your chance!

17.11.2024 18:01 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0

I don’t think it was - the episode came out so people bought, and Gwern confirmed it wasn’t him so people sold.

13.11.2024 21:45 β€” πŸ‘ 7    πŸ” 0    πŸ’¬ 1    πŸ“Œ 0
Post image

A tale of two platforms:

BlueSky user numbers have hit a new record high in recent days, while the number of people deleting their accounts on X/Twitter has rocketed πŸš€

13.11.2024 11:21 β€” πŸ‘ 20509    πŸ” 4125    πŸ’¬ 463    πŸ“Œ 417

Say 'I know all about this from a Lex Fridman interview I was listening to'.

13.11.2024 17:17 β€” πŸ‘ 16    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

:)

12.11.2024 22:08 β€” πŸ‘ 1    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

Nice, thanks!

12.11.2024 22:07 β€” πŸ‘ 0    πŸ” 0    πŸ’¬ 0    πŸ“Œ 0

@samatis is following 19 prominent accounts