@SashaGusevPosts profile picture
Sasha Gusev @SashaGusevPosts

Statistical geneticist | Associate Prof at @DanaFarber / @harvardmed / @DFCIPopSci | Blogging at theinfinitesimal.substack.co…

Joined January 2023
  • Posts 22,285
  • Following 2,909
  • Followers 22,067
3,656 Photos and videos
Pinned Post
Sasha Gusev
@SashaGusevPosts
21 Feb 2024
I've written about race, genetic ancestry, analyses of large biobanks, and human history gusevlab.org/projects/hsq/#h… I'll summarize the key points here 🧵:
49
308
1,332
590,365
When you post a chart that is already adjusted for inflation on X, the everything app
12
3
130
13,125
Sasha Gusev reposted
Marios Georgakis
@MariosGeorgakis
19h
Another common opinion, especially among computational researchers, is that a key barrier to integrating complex prediction models into clinical practice is that physicians are poor at interpreting risk estimates. People often underestimate how integral risk assessment is to routine clinical care and how often physicians need to communicate such risk to patients.
Sasha Gusev
@SashaGusevPosts
19h
There's lots of probabilistic thinking in medicine, it just gets encoded into risk groups that are tied to specific clinical actions (e.g. Framingham, CHADS-VASC, HAS-BLED, Gail/BCRAT, etc) or disease stage.
1
3
20
5,000
There's lots of probabilistic thinking in medicine, it just gets encoded into risk groups that are tied to specific clinical actions (e.g. Framingham, CHADS-VASC, HAS-BLED, Gail/BCRAT, etc) or disease stage.
Amanda Askell
@AmandaAskell
22h
Extracting a probability from a doctor is one of life's unnecessary boss battles. Even if you beg them for an interval-valued subjective probability, and at that point you're basically asking for their hunch. I don't know if they get sued for giving out information or something.
1
3
39
12,982
Some fun charts: 1. Each US generation is wealthier than the last.
180
213
1,762
617,655
10. Support for interracial marriage went from <50% to 94% in the past 30 years.
3
5
132
24,654
11. Overwhelming majorities of both parties believe morality, religion, success in life, criminality, and income are down to environment or choices rather than genes. /x
9
1
115
23,619
Very cool latent class analysis of national identity showing that ardent and restrictive views of nationalism in the US have declined after a post-9/11 high. Largely replaced with a "disengaged" group with little national identification.
3
5
44
5,098
Source: pixelsandpatterns.org/p/who-…

Who are the Americans?

On the 250th birthday of the United States, I reflect on thirty years of survey data to show a national identity in flux and make a bullish case for liberal nationalism.

pixelsandpatterns.org
6
997
Morris is pathologically incapable of distinguishing the criticism of a scholarly work from the criticism of a scholar. Then immediately pivots to attacking the scholar's academic roles and scolding the editors. It is the most pathetic form of cancel culture on full display.
Damien Morris
@DamienMorris
Jul 4
I don't think I've ever seen such a rhetorically manipulative abstract. Let's change the subject of the sentences to demonstrate how squalid this is: "Is Eric Turkheimer abhorrent or just controversial? To the extent that his work has been both limited in scientific value and potentially harmful, this paper asks the question: was his work abhorrent pseudoscience?" It's amazing that such a hatchet job can be published in an academic journal. And to think that Matthews sits on the BGA Public Science Committee which presumes to tell the rest of the behavioural genetics community what is appropriate to say and think. Shame on Matthews and Turkheimer for writing this and shame on Behavior Genetics for publishing it.
6
5
59
9,265
I do find this behavior useful in one sense though. Students regularly ask me for advice on how to engage thoughtfully on social media. My advice: don't act like this guy!
1
1
25
1,124
Sasha Gusev reposted
BuccoCapital Bloke
@buccocapital
Jul 4
Occasionally I’ll read something so idiotic only to realize I’ve been forced into the For You feed. It’s so bad. It’s insulting. It’s disgusting and pathetic. It makes me feel ashamed for being on this dogshit app. Honestly impressive how bad it is. Terrible, terrible product
47
47
1,035
172,462
Hmmm ...
Epoch AI
@EpochAIResearch
Jul 2
Introducing EBR-bench, our new benchmark to measure on-the-fly learning. AI repeatedly plays a challenging board game called Earthborne Rangers and tries to learn from its mistakes. So far: no signs of improvement.
1
13
5,125
Oh .
1
8
1,293
Load more