Varun
LightNews LightNews
varunr.bsky.social
Varun
@varunr.bsky.social
330 followers 480 following 280 posts
Data science at Slack. Views are my own (obviously). Largely data posting and politics re-posting.

Politics blog: https://a-humble-politics-blog.ghost.io/
Data science + engineering blog: https://varunprajan.github.io/
Posts Replies Media Videos
varunr.bsky.social
lol
Jeff Yass claims prediction markets could have prevented the Iraq war
November 23, 2025 at 11:15 PM
varunr.bsky.social
The other important reason that charts might lead us astray is that metrics are rarely so straightforward as “deaths”. This is where data science and quantitative methodology can be genuinely useful — not so much in building charts, but in helping to understand what they really mean. There is a “third way” that’s better than either of the first two ways. We don’t have to be numerically illiterate or reactionary against quantitative information, like so many on social media. Yet neither do we have to be the kind of supercilious asshole who thinks that a chart represents a knockout blow (like so many others on social media). Instead of anecdotes, on the one hand, or scientism, on the other, we can do (data) science.
November 16, 2025 at 11:11 PM
varunr.bsky.social
tough to tell who the worst edblog writer for TPM is
November 12, 2025 at 6:17 PM
varunr.bsky.social
lol incredible
November 12, 2025 at 1:24 PM
varunr.bsky.social
"Sam Altman says the one-person billion-dollar company is coming. Maybe I could be that person—if only I could get my colleagues to shut up and stop lying."

www.wired.com/story/all-my...
November 12, 2025 at 1:23 PM
varunr.bsky.social
I imagine he knows that already, since they've been saying it to his face for years now. The pull of making money is too powerful
"AnnCoulter told me flat-out to my face that she couldn't vote for me "because you're an Indian"" "Apoorva Ramaswamy on comment about husband's dark skin: "Not much we can do about that one""
November 11, 2025 at 12:54 PM
varunr.bsky.social
bit off topic, but I feel like this summarizes the right wing's position on free speech pretty well (from David Hamburger Helper's blog)
David Hamburger Helper's blog: "Calling someone a "nazi" is a permission slip for violence" David Hamburger Helper's blog: "Words are not violence"
November 11, 2025 at 12:41 PM
varunr.bsky.social
November 10, 2025 at 2:37 AM
varunr.bsky.social
Of course they haven't arrested the guy, they probably agree with him
November 6, 2025 at 10:30 PM
varunr.bsky.social
loved this section on AtlasIntel
AtlasIntel's final poll was off by 11 points
November 5, 2025 at 1:24 PM
varunr.bsky.social
from the cuomo watch party lol
November 5, 2025 at 3:11 AM
varunr.bsky.social
what are these people talking about?
"A little-noted element propelled mamdani's rise: gen z loneliness" nyt headline
November 5, 2025 at 2:42 AM
varunr.bsky.social
update
November 5, 2025 at 2:39 AM
varunr.bsky.social
lol at the TV screens
Cuomo election night watch party watching TV showing Cuomo losing
November 5, 2025 at 2:36 AM
varunr.bsky.social
Grokipedia uses "Latinx". Woke much?
November 2, 2025 at 10:40 PM
varunr.bsky.social
Ro Khanna and Zohran Mamadani appear together at a South Asian/Muslim rally in Jamaica NY
November 2, 2025 at 10:20 PM
varunr.bsky.social
have these people been awake the last 8 years?
Ruth Marcus: This court would never "rewrite the clear language of the Constitution" Michael Luo: "It's almost like he's testing the limits of how far he can go." Yeah, almost Ruth Marcus: "I think the door's pretty firmly shut" on Trump running for a 3rd term
November 2, 2025 at 10:17 PM
varunr.bsky.social
patiently waiting for this to spread to tech companies
October 30, 2025 at 7:31 PM
varunr.bsky.social
helluva guy
Cuomo made 4.9M last year, and won't tell anyone who was paying him
October 30, 2025 at 1:51 PM
varunr.bsky.social
I assume this is separate from the "Hyperion" project Matt Levine was talking about yesterday? I thought these companies didn't want debt on their books
Matt Levine talking about the Hyperion SPV
October 30, 2025 at 1:40 PM
varunr.bsky.social
incredible (sorry I'm not linking to the NY Post)
The Times of London retracted an interview with Bill de Blasio in which he criticized Mamdani after discovering that the person they spoke with was not, in fact, de Blasio.
October 29, 2025 at 12:57 PM
varunr.bsky.social
good example of the difference between relative and absolute metrics www.wired.com/story/chatgp...
In a given week, OpenAI estimated that around 0.07 percent of active ChatGPT users show “possible signs of mental health emergencies related to psychosis or mania” and 0.15 percent “have conversations that include explicit indicators of potential suicidal planning or intent.” OpenAI also looked at the share of ChatGPT users who appear to be overly emotionally reliant on the chatbot “at the expense of real-world relationships, their well-being, or obligations.” It found that about 0.15 percent of active users exhibit behavior that indicates potential “heightened levels” of emotional attachment to ChatGPT weekly. The company cautions that these messages can be difficult to detect and measure given how relatively rare they are, and there could be some overlap between the three categories. OpenAI CEO Sam Altman said earlier this month that ChatGPT now has 800 million weekly active users. The company’s estimates therefore suggest that every seven days, around 560,000 people may be exchanging messages with ChatGPT that indicate they are experiencing mania or psychosis. About 2.4 million more are possibly expressing suicidal ideations or prioritizing talking to ChatGPT over their loved ones, school, or work.
October 27, 2025 at 10:58 PM
varunr.bsky.social
let's go
Early voted for Zohran
October 25, 2025 at 3:12 PM
varunr.bsky.social
Pretty racist passage in this review of "What is War" by (of course) the NYT.

Otake has lived in NYC since 1976! She speaks English with an accent, but it's perfectly comprehensible
"This is potent material, but “What Is War” doesn’t quite ignite it. A big part of the problem is language. Otake’s and Hui’s English is sometimes hard to understand, and their lack of fluency makes the speeches tentative and flat. Otake inflects the speeches, both hers and Hui’s, with her always expressive body language — clawing at herself, crumpling — but this serves to highlight the verbal weakness."
October 25, 2025 at 2:28 PM
varunr.bsky.social
LeCun back in April: "This is most definitely not the death of FAIR

fortune.com/2025/04/10/m...
FAIR—an acronym for Fundamental AI Research—was once the crown jewel of AI development at Meta. But as Mark Zuckerberg has pivoted the company toward generative AI products over the past two years, the vaunted lab has become something of an orphan inside the organization, increasingly shoved out of the limelight by more commercially focused AI groups within the company. The newest Llama model, for instance, was the product of Meta’s separate GenAI team, not FAIR. Meanwhile, FAIR has languished, with talented researchers departing for rival companies and startups: More than half of the 14 authors of the original Llama research paper published in February 2023 had left the company six months later, while at least eight top researchers have left over the past year. The lab has been “dying a slow death,” according to one former Meta employee who spoke to Fortune over the past week, echoing a sentiment from other ex–FAIR team members. Pineau’s departure, in this view, can be interpreted as the lab’s death rattle. Yann LeCun, Meta’s chief AI scientist, who is considered one of the “godfathers” of deep learning and who founded FAIR, is now leading the team temporarily as Meta does an external search to replace Pineau (who is staying on until May). He denies FAIR is fading into oblivion, telling Fortune by email that “this is most definitely not the death of FAIR.” To the contrary, he insists the lab is about to experience a new dawn, and the existence of the GenAI product group actually enables FAIR to refocus on longer-term AI research.
October 22, 2025 at 12:30 PM