Arnav Das
LightNews LightNews
banner
arnav7.bsky.social
Arnav Das
@arnav7.bsky.social
1.7K followers 12K following 820 posts
I try to "Observe, Reflect and Explore around" in different ways

My posts here will be a mishmash of personal opinions and smol curious excursions..
Posts Replies Media Videos
Pinned
arnav7.bsky.social
Arnav Das @arnav7.bsky.social · Oct 19
Ok since pinning has finally been added to this place,

This is a reference post of all my excursions I try around..
Reposted by Arnav Das
markrubin.bsky.social
"Much of what we know comes from what other people tell us. Thus, the epistemology of testimony is a critical area of social epistemology."

"Liars and deceivers are what we might call epistemic adversaries."
December 9, 2025 at 6:52 AM
Reposted by Arnav Das
drlpretorius.bsky.social
Another big celebration moment for 2025! 🎉🎉 Thanks to the students who nominated me for the MGA Supervisor of the Year! It is an honour to help guide you through your PhDs to becoming independent scholars!
Photo of four smiling people holding several bunches of flowers and a framed certificate A photo of a framed certificate. The text says that it is a Special Commendation for Supervisor of the Year presented to Dr Lynette Pretorius Photo of a framed certificate in front of several flowers
December 9, 2025 at 7:14 AM
Reposted by Arnav Das
wiringthebrain.bsky.social
Free will, the genomic code, and high-tech eugenics open.spotify.com/episode/4IRt...
#1132 Kevin Mitchell: Free Will, Robert Sapolsky, the Genomic Code, and Human Embryo Editing
open.spotify.com
December 9, 2025 at 7:42 AM
Reposted by Arnav Das
dingdingpeng.the100.ci
No take only throw meme. First panel: dog with a frisbee asking: Pls great power? Second panel: dog angry because somebody tries to take frisbee: No great responsibility! Third panel: angry do close-up: ONLY POWER
December 9, 2025 at 8:04 AM
Reposted by Arnav Das
koenfucius.bsky.social
Has @anilseth.bsky.social identified an underrated reason why “AI consciousness” isn’t imminent?

Humans—unlike AI—never get stuck in infinite loops.

Our existence is rooted in time and progressing entropy. Both may be consciousness prerequisites:

buff.ly/8PxbtRs
December 8, 2025 at 11:35 AM
Reposted by Arnav Das
markrubin.bsky.social
“Positivism Creep”

“The subtle, often unacknowledged infiltration of positivist or post-positivist assumptions into methodologies where they do not naturally, philosophically, practically, or epistemologically belong.”

By @thomasgraves.bsky.social @maddipow.bsky.social @annayahprosser.bsky.social
Many developments to reform the research landscape have occurred over the past decade. These changes have been made with broad goals to improve the ‘openness’ of research and often assumed to be ‘methodologically -agnostic’; that is, they ostensibly have benefits for all researchers occupying all epistemological and methodological positions. ‘Open science’ initiatives such as study pre-registration (i.e., specifying research aims and analytical plan ahead of data access), open data sharing, open-access publication, and open materials sharing are becoming increasingly mainstream across many fields within social research and the natural sciences. While there has been much criticism of these interventions, largely from the qualitative research community, we want to draw attention to a troubling trend in the promotion of open science: the leaking of standards relevant only to quantitative research to all paradigms. Or, as others refer to it, “positivism creep”. Here, we situate positivism creep (i.e., the creeping of positivist conventions to all research) within research policy, we highlight its increasing prevalence within open science reforms, and we warn against a future which could alienate many non-positivist scholars. We argue that the primary framing of open science as the pursuit of reproducibility and objectivity risks promoting positivism creep in the social sciences and humanities. In particular, we suggest that overly strict open research requirements placed by funders may reduce the range and variety of epistemological positions that can be taken by researchers, with particularly deleterious effects for qualitative researchers.
December 5, 2025 at 7:07 PM
Reposted by Arnav Das
emp1.bsky.social
Hebbian unsupervised learning updates weights locally based on correlations in neural activity. This research advances biologically plausible convolutional neural networks (CNNs) by showing that Hard Winner-Takes-All (WTA) competition can achieve performance comparable to backpropagation.
December 7, 2025 at 10:39 PM
Reposted by Arnav Das
lisabortolotti.com
For research updates, new books, interviews with experts, spotlights on projects, and conference reports, follow @imperfectcognitions.com. A blog at the intersection of #philosophy, #psychology, and #mentalhealth: a new post every week, rain or shine. #philosophyforeveryone #philsky
Two heads, one green and one yellow, and an orange background.
December 7, 2025 at 7:52 PM
Reposted by Arnav Das
djnavarro.net
Ah fuck it. In an expression of my frustration at having been unable to focus on things I actually *want* to talk about, I wrapped up this little blog post about managing git/github configurations for multiple identities on the one computer. As always, read it or don't :)
Two githubs, one laptop – Notes from a data witch
Some notes on managing multiple sets of github credentials on the one machine
blog.djnavarro.net
December 7, 2025 at 9:46 PM
Reposted by Arnav Das
grimalkina.bsky.social
be a person willing to dig deep into methods and you will defang a lot of bullshit
iansociologo.bsky.social Ian Carrillo @iansociologo.bsky.social · 1d
Excellent article. Paragraph after paragraph is damning.
December 7, 2025 at 6:41 PM
Reposted by Arnav Das
seanmcarroll.bsky.social
Immaturity masquerading as strength is the defining personal characteristic of our age.
gilduran.com Gil Durán @gilduran.com · 2d
A Palantir billionaire just called for public hangings.

“It's time to bring back masculine leadership to protect our most vulnerable,” writes Joe Lonsdale.

This is where Silicon Valley authoritarianism is heading:

www.thenerdreich.com/joe-lonsdale...
Joe Lonsdale Calls For Public Hangings
Silicon Valley radicalization escalates
www.thenerdreich.com
December 7, 2025 at 4:59 PM
Reposted by Arnav Das
vilgothuhn.bsky.social
I wrote a blogpost on how I don't think Lindley's paradox is a good justification for the practice of "p-value circling" (treating significant p-values close to 0.05 with suspicion). Also some other stuff. #rstats #stats
I don’t think Lindley’s paradox supports p-circling – Vilgot's website
Don’t give p-values a role they’re not made for
vilgot-huhn.github.io
December 7, 2025 at 1:12 PM
Reposted by Arnav Das
koenfucius.bsky.social
Some behaviours can easily be turned into habits—just repeat until it sticks.

Others are much harder.

Why the difference?

psyche.co/ideas/why-so...
December 7, 2025 at 1:03 PM
Reposted by Arnav Das
koenfucius.bsky.social
New research suggests fluid intelligence—the capacity to solve new problems and recognize patterns—is associated with the ability to map objects or items of information, capturing how they relate to each other in space or conceptually:

buff.ly/DHoRjsv
via @psypost.bsky.social
December 7, 2025 at 7:53 AM
Reposted by Arnav Das
koenfucius.bsky.social
How do you measure someone’s ‘moral goodness’?

With great difficulty, argue @jessiesun.bsky.social & @eschwitz.bsky.social—it’s not even certain there is such a general trait, and the usual measures (eg self-reports, behavioural observation) are problematic:
buff.ly/a9jgI75
HT @bxjaeger.bsky.social
December 7, 2025 at 9:18 AM
Reposted by Arnav Das
bayesianboy.bsky.social
What strikes me over and over again about the reality we are currently creating with so-called “generative AI” is that it’s always pitched as saving human time and effort. But this is only true if you are a parasitic grifter. For anyone with principles, anyone who cares about reality
cfiesler.bsky.social Dr. Casey Fiesler @cfiesler.bsky.social · 3d
I mean we are absolutely in a place now where the only solution to this information disorder is for everyone to constantly evaluate the source of information. Never trust a chatbot, but also don't believe a video unless you know and trust where it comes from.

Unfortunately... that's a lot of work.
December 6, 2025 at 1:22 AM
Reposted by Arnav Das
This guy hates sociology.
statmodeling.stat.columbia.edu/2025/12/06/t...
This guy hates sociology. | Statistical Modeling, Causal Inference, and Social Science
statmodeling.stat.columbia.edu
December 7, 2025 at 1:51 PM
Reposted by Arnav Das
robsica.bsky.social
"In evolutionary terms, the algorithmic boss is a ghost alpha: powerful but bodiless. And our brains don’t quite know what to do with that."
Why AI Will Never Be Your Leader
Personal Perspective: We’re letting algorithms increasingly guide our choices. AI can manage tasks better than any human, yet it can’t earn trust, loyalty, or moral authority.
www.psychologytoday.com
December 7, 2025 at 12:38 PM
Reposted by Arnav Das
chriswolfvision.bsky.social
Just finished "Blindsight" by Peter Watts. It's hard fiction and first contact with lots of questions on consciousness and empathy, which is my favorite aspect of the book.

Not a fan of having a vampire in SciFi, and I found the last 40 or so pages a bit confusing, but I still recommend it.

1/3
December 7, 2025 at 12:45 PM
Reposted by Arnav Das
djnavarro.net
you know, i'm starting to think that the UK might have a bit of a transphobia problem

www.bbc.com/news/article...
Trans women will not be allowed to attend main Labour Women's Conference events
The conference will take place in 2026 after being cancelled for a year in light of the Supreme Court's gender ruling.
www.bbc.com
December 6, 2025 at 6:25 PM
Reposted by Arnav Das
koenfucius.bsky.social
The momentary lapses of attention we experience after a sleep deprived night is when the brain catches up on the maintenance and flushing out it would otherwise have done during proper sleep, explains @halmacmc:

buff.ly/O9blMv1
December 6, 2025 at 7:51 PM
Reposted by Arnav Das
mewtation.bsky.social
This thread has been fascinating to think about re: my last workplace, which kept trying to cultivate a learning culture while also constantly undermining it with the real priorities
grimalkina.bsky.social Cat Hicks @grimalkina.bsky.social · 2d
Miller et al. document e.g., that *unrealistic productivity expectations* complete change how someone interprets the meaning of the tool, workplace, and role expectation interaction. This is exactly what I predicted too. We keep converging on this.

arxiv.org/abs/2507.21280
"Maybe We Need Some More Examples:" Individual and Team Drivers of Developer GenAI Tool Use
Despite the widespread availability of generative AI tools in software engineering, developer adoption remains uneven. This unevenness is problematic because it hampers productivity efforts, frustrate...
arxiv.org
December 6, 2025 at 8:13 PM
Reposted by Arnav Das
vikramsaraph.com
uiua has got to be the weirdest programming language I’ve ever seen: www.uiua.org
Uiua
A tacit array programming language
www.uiua.org
December 6, 2025 at 9:54 PM
Reposted by Arnav Das
robsica.bsky.social
"Reason is a metarepresentational mechanism... If many, or maybe all, metarepresentational mechanisms serve social functions—even if the functions are not exactly the same—this would further argue for their more integrated study"
Reasoning and Argumentation
oecs.mit.edu
December 7, 2025 at 3:24 AM
Reposted by Arnav Das
olivia.science
Is there any realistic alternative because the profiles with pdfs and whatever are so useful? Shall we build one?
December 5, 2025 at 6:25 PM