Deena Mousa

719 posts

Deena Mousa banner
Deena Mousa

Deena Mousa

@deenamousa

Research & grantmaking @coeff_giving • Leave me anonymous feedback: https://t.co/0YmbLhC2dj

New York Katılım Kasım 2016
538 Takip Edilen2.5K Takipçiler
Sabitlenmiş Tweet
Deena Mousa
Deena Mousa@deenamousa·
Roundup of my favorite May reading: India's IT revenue and employment decoupled for the first time in 2 decades, ChatGPT's obsession with goblins is actually very human, and in countries with fewer than 1 doctor/10,000 people, the AI-and-jobs debate looks a *little* different
Deena Mousa tweet media
English
1
2
10
589
Deena Mousa
Deena Mousa@deenamousa·
Roundup of my favorite May reading: India's IT revenue and employment decoupled for the first time in 2 decades, ChatGPT's obsession with goblins is actually very human, and in countries with fewer than 1 doctor/10,000 people, the AI-and-jobs debate looks a *little* different
Deena Mousa tweet media
English
1
2
10
589
Deena Mousa
Deena Mousa@deenamousa·
@DKThomp The window during which AI is sufficiently bad at video for this to be the moat is likely very short
English
0
0
5
299
Derek Thompson
Derek Thompson@DKThomp·
i don't want to be a doomer about this but AI is going to eat more of the world of writing unless either: (a) we get much better at recognizing AI writing (kind of doubtful, considering the tech will never be worse than it is now at mimicking the voice of specific writers); or (b) the social penalty for writing with AI goes up and ppl feel more shame about it (also kind of doubtful, bc of [a]) one effect this could have is to reduce trust in more forms of writing, which further shifts demand toward orality and video, where AI is still a bit uncanny valley-ish, and everything keeps becoming television x.com/TomRachman/sta…
English
19
8
128
21.9K
Derek Thompson
Derek Thompson@DKThomp·
that both of these stories were published on the same day is almost too convenient for think-piece lede-writing purposes
Derek Thompson tweet mediaDerek Thompson tweet media
English
3
15
172
35.8K
Deena Mousa
Deena Mousa@deenamousa·
@rSanti97 I heard it has its own beige microsite actually
English
0
0
22
567
Santi Ruiz
Santi Ruiz@rSanti97·
Have you seen the new AI safety paper? It's on Guidelight. It's literally on Forethought. You can probably find it on Foresight. Dude it's at Coefficient. Just go to the Future of Life. It's hosted at Lighthaven. It's on Longview. Just open Conjecture. It's at Elicit.
English
24
26
458
15.3K
Deena Mousa
Deena Mousa@deenamousa·
@StefanFSchubert The potential economic effects are ambiguous enough, and have the potential to hit differently sectors enough, that this doesn’t feel satisfying. It’s important for policy to have some way of forming a belief about what the labor market and work will look like specifically.
English
0
0
0
79
Derek Thompson
Derek Thompson@DKThomp·
appreciate irony as much as the next guy but writing a book called The Future of Truth that uses artificial intelligence to insert a false quote from a real book called Artificial Unintelligence is really putting in on thick
Derek Thompson tweet mediaDerek Thompson tweet media
English
13
81
500
45.5K
Deena Mousa
Deena Mousa@deenamousa·
I have a bone to pick with AI usage data and AI exposure indexes:
Deena Mousa tweet media
English
1
0
4
397
Deena Mousa
Deena Mousa@deenamousa·
New post: Unfortunately, nobody knows what AI exposure means.
Deena Mousa tweet media
English
2
6
33
3.9K
Deena Mousa
Deena Mousa@deenamousa·
@AlexanderMcCoy4 Neither! Just that, especially given uncertainty around AI timeline and impacts, it’s important to note whether the thing you suggest is just generally good policy vs “specific thing you should do given AI will implicitly do X but is otherwise bad or unimportant to do”
English
1
0
1
24
Alexander McCoy
Alexander McCoy@AlexanderMcCoy4·
@deenamousa To be clear and make sure I’m interpreting you correctly, are you saying that it is Good or Bad if the proposed solution is consistent with what someone would be advocating for in a counterfactual world without AI?
English
1
0
1
57
Deena Mousa
Deena Mousa@deenamousa·
My first question when it comes to suggestions about how to respond to AI via policy is often: 'How is this different from what you'd suggest doing if AI didn't exist?' It does change prioritization, but being clear about when that's the case matters
Nat Purser@NatPurser

imo, people overindex on addressing ai’s impacts through novel, bespoke solutions, when the best proposals are often normal ideas with historical precedent — modernizing unemployment insurance, improving K-12 education, raising taxes to finance welfare state expansion, etc.

English
2
0
7
859

Keşfet

@joshgans @danbjork @davideoks @carolyn_sms @RyanReedHill @DKThomp @rSanti97 @StefanFSchubert