Dean Eckles
LightNews LightNews
banner
eckles.bsky.social
Dean Eckles
@eckles.bsky.social
9.5K followers 1.3K following 350 posts

networks, contagion, causality
faculty at MIT

Videos
Mathematics 19%
Physics 19%
Posts Replies Media Videos

Reposted by Dean Eckles

ianhussey.mmmdata.io
Ian Hussey @ianhussey.mmmdata.io · 5h
> "There is no suggestion of malpractice"

The article's pubpeer thread would strongly disagree.

pubpeer.com/publications...

eckles.bsky.social
Dean Eckles @eckles.bsky.social · 14h
I guess the differences between the outcome being y_post-y_pre and y_post are mainly finite sample ones depending on the choice of the smoother. So in reasonable samples those curves should be similar either way.

Or what do you have in mind here?

Reposted by Will Jennings, Caroline Krafft, Alan Richardson , and 4 more

elliottpayne.org
Elliott Payne @elliottpayne.org · 1d
ICE just indiscriminately stopped a citizen at a bus stop in front of the co-op. If this is how they treat the President of the Minneapolis City Council who is legally observing, just think of what they’re doing to everyone else in our City.

Reposted by Dean Eckles

simonwillison.net
Simon Willison @simonwillison.net · 1d
Wrote up my first impressions of Claude Cowork, Anthropic's new general purpose agent released today for $100+/month subscribers as part of their macOS desktop app simonwillison.net/2026/Jan/12/...
First impressions of Claude Cowork, Anthropic’s general agent
New from Anthropic today is Claude Cowork, a “research preview” that they describe as “Claude Code for the rest of your work”. It’s currently available only to Max subscribers ($100 …
simonwillison.net

Reposted by Aaron Sojourner

eckles.bsky.social
Dean Eckles @eckles.bsky.social · 1d
Working on implications of AI for work and the economy?

@professorsinan.bsky.social @dholtz.bsky.social and our team at the MIT IDE are organizing this conference. Submit abstracts by Friday, Jan 16.

ide.mit.edu/events/2026-...
2026 The Business Implications of Generative AI @ MIT (BIG.AI@MIT) - MIT Initiative on the Digital Economy
02/04/2026 - 03/04/2026 @ 8:00 am - 6:00 pm -
ide.mit.edu

Reposted by Dean Eckles, Jan R. Boehnke

eikofried.bsky.social
Eiko Fried @eikofried.bsky.social · 2d
Our machine learning competition on forecasting depression is online! We'd love for as many people as possible to participate. Please share in your respective networks — thank you :).

www.staff.universiteitleiden.nl/announcement...
Machine-learning competition on forecasting depression in young adults - Leiden University
After nearly 5 years of work, the research team led by Eiko Fried has finished data collection in the ERC-funded WARN-D project on building a personalised early warning system for depression. But just...
www.staff.universiteitleiden.nl

Reposted by Brendan Nyhan, Robert W. Wallace, Rosemary A. Joyce , and 16 more

bgrueskin.bsky.social
Bill Grueskin @bgrueskin.bsky.social · 2d
Powell, showing more guts than almost any GOP legislators:

"The threat of criminal charges is a consequence of the Fed setting interest rates based on our best assessment of what will serve the public, rather than following the preferences of the President."

www.federalreserve.gov/newsevents/s...

Reposted by Anne Applebaum, Stephen M. Walt, Kori Schake , and 88 more

federalreserve.gov
Federal Reserve @federalreserve.gov · 2d
Video message from Federal Reserve Chair Jerome H. Powell:
www.youtube.com/watch?v=KckG...
www.federalreserve.gov/newsevents/s...
Statement by Federal Reserve Chair Jerome H. Powell
YouTube video by Federal Reserve
www.youtube.com

eckles.bsky.social
Dean Eckles @eckles.bsky.social · 2d
That's not what this says

Reposted by Dean Eckles

rtodkelly.bsky.social
Schrödinger's Sneetch Belly @rtodkelly.bsky.social · 2d
Grifters always assume everyone everywhere is grifting.

Reposted by Melanie C. Green, Michael D. McDonald, Anne Norton , and 1 more

atrupar.com
Aaron Rupar @atrupar.com · 2d
WELKER: Where is that $50,000? Did you keep it?

HOMAN: I didn't take $50k from anybody & that's a question for the FBI. Bottom line I did nothing illegal.

WELKER: But was there $50k in the bag?

HOMAN: I'm not addressing it

WELKER: But did you keep the money?

HOMAN: $50k is ridiculous

Reposted by Dean Eckles

natematias.bsky.social
J. Nathan Matias @natematias.bsky.social · 3d
Proposal: call reverse chronological algorithms "spew machines" since they reward whoever is most extremely online and posts the most.

I think that might counteract some of the weird nostalgia that leads so many to think it's actually some kind of pure symbol of the halycon Internet.

Reposted by Rebecca Solnit, Steve Peers, Dean Eckles , and 2 more

onestpress.onestnetwork.com
Olga Nesterova @onestpress.onestnetwork.com · 3d
And despite it all, this is Tehran 🇮🇷

Reposted by Nandita Sharma, Will Jennings, Brendan Nyhan , and 13 more

donmoyn.bsky.social
Don Moynihan @donmoyn.bsky.social · 3d
This is the second time in three months a DHS agent called a US citizen a "bitch" as he shot her. donmoynihan.substack.com/p/whos-threa...
The video also allegedly shows the CBP agent saying “Do something, bitch” before shots were fired. In private Signal chats, the CBP agent boasted to other officers: “I fired 5 rounds and she had 7 holes. Put that in your book boys.” If he felt comfortable enough communicating that to fellow employees, it speaks to an irredeemable organizational culture.

Reposted by Lisa W. Fazio, David R. Miller, Dean Eckles

brendannyhan.bsky.social
Brendan Nyhan @brendannyhan.bsky.social · 3d
We should expect more from government than being "not crazy" and not write columns praising them for meeting this pathetically low bar statmodeling.stat.columbia.edu/2026/01/08/t...

Reposted by Dean Eckles

jeffreybigham.com
Jeffrey P. Bigham @jeffreybigham.com · 3d
i feel like it's somehow a very controversial take to say one generally supports immigration enforcement but not shooting unarmed women in the face

Reposted by Dean Eckles

walterolson.bsky.social
Walter Olson @walterolson.bsky.social · 3d
IJ files comments: DHS’s sweeping proposal to compel DNA and biometrics submissions from citizens "is exactly the kind of generalized, future-facing data collection that the
Fourth Amendment is meant to guard against" and has never been authorized by Congress. ij.org/wp-content/u...

eckles.bsky.social
Dean Eckles @eckles.bsky.social · 4d
The last digits are also kind of confusing here! Like how many degrees of longitude wide is Cambridge?

eckles.bsky.social
Dean Eckles @eckles.bsky.social · 4d
Thanks Gemini!

eckles.bsky.social
Dean Eckles @eckles.bsky.social · 4d
I guess this is one way to get Wall Street people to join Truth Social
www.ft.com/content/7ea9...
If a jobs report leaks on a social media platform no one uses, does it move the market?
Data forecasting made easy
www.ft.com

eckles.bsky.social
Dean Eckles @eckles.bsky.social · 4d
I guess some of this kind of work just operates in a finite population / deterministic potential outcome setup. So then for a given draw, once you simulate the potential outcomes, they fully define the truth.

Reposted by Scott A. Imberman, Dean Eckles

bencasselman.bsky.social
Ben Casselman @bencasselman.bsky.social · 4d
This is pretty wild: Health care & social assistance accounted for essentially ALL of private-sector job growth in 2025 (713k out of the 733k private-sector job gains).

Reposted by Gordon Pennycook, Dean Eckles

tomerullman.bsky.social
Tomer Ullman @tomerullman.bsky.social · 5d
once again being driven insane by ML conference submissions

eckles.bsky.social
Dean Eckles @eckles.bsky.social · 5d
Makes me think about how this shows up in ideal point models of legislators
andycraig.bsky.social Andy Craig @andycraig.bsky.social · 5d
There's no rhyme or reason to which Dems do this on any given nominee, and they never explain, because it really is as stupid as they do it to game metrics so they can't be tagged "most liberal." It's not because Republicans care, they don't need them. It's not a tactical maneuver. It's just dumb.
paleofuture.bsky.social Matt Novak @paleofuture.bsky.social · 5d
Amy Klobuchar voted for one of Trump's nominees the day after a woman in her state was murdered by ICE.

Reposted by Dean Eckles

andycraig.bsky.social
Andy Craig @andycraig.bsky.social · 5d
There's no rhyme or reason to which Dems do this on any given nominee, and they never explain, because it really is as stupid as they do it to game metrics so they can't be tagged "most liberal." It's not because Republicans care, they don't need them. It's not a tactical maneuver. It's just dumb.
paleofuture.bsky.social Matt Novak @paleofuture.bsky.social · 5d
Amy Klobuchar voted for one of Trump's nominees the day after a woman in her state was murdered by ICE.
fascistdemtracker.bsky.social Fascist Dem Tracker @fascistdemtracker.bsky.social · 5d
Alexander Van Hook is confirmed 53-40 to be a District Judge of Western Louisiana. According to @senatepress.bsky.social, the Dems yeas are Dick Durbin (IL), Maggie Hassan (NH), Tim Kaine (VA), Amy Klobuchar (MN), Jeanne Shaheen (NH), and Peter Welch (VT). Angus King (I-ME) was also a yea.

Reposted by Dorothy Bishop, Martin Tomko, Dean Eckles

jackfitzgerald.bsky.social
Jack Fitzgerald @jackfitzgerald.bsky.social · 5d
My Matters Arising concerning a paper on the legal determinants of terrorism is now out in @nathumbehav.nature.com. The original paper is now retracted. To learn why, read on for a story of irregularities, imputations, and impossible values. 1/x

doi.org/10.1038/s41562-025-02347-7
Client Challenge
doi.org

eckles.bsky.social
Dean Eckles @eckles.bsky.social · 6d
Also, when I've encountered this thinking, it is perhaps as much about the absurd incentives/goals for MDs in academic medical centers as it is competence. Though then it is a reason to basically not trust papers where all lead/senior authors are MDs, even if they brought eg some statistician along.

eckles.bsky.social
Dean Eckles @eckles.bsky.social · 6d
Don't know how many MDs you know, but maybe that still fits with the heuristic doing well on average?

I would be interested in overall stats about how much of the literature has only MD PIs.

eckles.bsky.social
Dean Eckles @eckles.bsky.social · 6d
"If a quantitative paper's authors are all MDs, it is usually worth ignoring."

Have people heard this heuristic? If so, where?

Reposted by Peter Thorne, Rosemary A. Joyce, David R. Miller , and 4 more

donmoyn.bsky.social
Don Moynihan @donmoyn.bsky.social · 6d
"Was the officer struck by the vehicle, as President Trump insists, or did the car pass by or around him?"
If we only had some trusted institution that could adjudicate between competing claims by relying on facts.
The basic sequence of the ICE shooting in Minneapolis on Wednesday is not in dispute: A Honda Pilot reversed, then moved forward; an ICE agent was close to the vehicle; the agent fired shots. A 37-year-old woman was killed. But that sequence has been interpreted in radically different ways. Was the officer struck by the vehicle, as President Trump insists, or did the car pass by or around him? Was he positioned in front of the vehicle or to the side? Did he have a genuine, reasonable fear for his life in that moment, or did he create the very danger he then used lethal force to escape?