Lightnews — Scholar-powered news

Ben Golub @bengolub.bsky.social · 1d

amazing tweet

4

Ben Golub @bengolub.bsky.social · 1d

presented without comment

3 15

Reposted by Ben Golub

Julian Reif @julianreif.bsky.social · 8d

I tried this on a manuscript—I’m impressed! It flagged:

- A missing variable in a first-order condition
- An outdated manual calc (“drop from X to Y is not Z%”)
- Inconsistent notation
- Miscellaneous typos (“and and”, “+ +”)
- Missing subscripts
- Several unclear passages

Ben Golub @bengolub.bsky.social · Jul 24

I've been working on a new tool, Refine, to make scholars more productive. If you're interested in being among the very first to try the beta, please read on.

Refine leverages the best current AI models to draw your attention to potential errors and clarity issues in research paper drafts.

1/

1 4 18

Ben Golub @bengolub.bsky.social · 14d

We'll make one. Email us at [email protected] or dm me so we have your info

1

Ben Golub @bengolub.bsky.social · 14d

Good idea! We'll try to develop something like that. Departments can get student subscriptions at very good rates.

1 1

Ben Golub @bengolub.bsky.social · 14d

You might have seen me posting about it before. One new thing is that we now give a free preview automatically to any new user.

If you signed up before and would like a free preview, just send a message to [email protected].

2

Ben Golub @bengolub.bsky.social · 14d

This is the kind of feedback it gives -

We've been using it mostly on technical manuscripts in social science, statistics, applied math, philosophy, and other topics. Would welcome feedback on how it performs on your research.

1 4

Reposted by Ben Golub

Florian Hollenbach @fhollenbach.org · Sep 5

I have just used it for the first time and let me tell you, it is well worth the money. Highly recommended for a last check before submission. Great stuff 👏

Ben Golub @bengolub.bsky.social · Jul 24

I've been working on a new tool, Refine, to make scholars more productive. If you're interested in being among the very first to try the beta, please read on.

Refine leverages the best current AI models to draw your attention to potential errors and clarity issues in research paper drafts.

1/

2 6

Reposted by Ben Golub

Alex Engler @alexengler.bsky.social · Aug 31

This @beijingpalmer.bsky.social post reminded me I have been meaning to update @pewresearch.org's estimate of total newspaper newsroom jobs - which was 30,820 as of 2020

Using BLS OEWS data (w/ same methodology as Pew) the 2024 number is 29,260 - down another 5%

www.pewresearch.org/chart/sotnm-...

1 15 34

Reposted by Ben Golub

Ryan Moulton @moultano.bsky.social · Aug 30

Survivorship bias maybe. There were other chesses.

🐔 Brian Bucklew 🐔 ₑͤ＞∿＜ₑͤ ∞🌮 @unormal.bsky.social · Aug 29

chess is a fucked up game it really is way more finely balanced than it has any right to be

3 1 24

Reposted by Ben Golub

Tim Bresnahan @timobres.bsky.social · Aug 5

Peter Temin, economic historian at MIT, has passed away. He worked both on problems of understanding the past (e.g. "Did Monetary Forces Cause the Great Depression?") and how how the past shapes the present (e.g. "The Vanishing Middle Class.") Terrific scholar and mentor, generous critic.

3 19 83

Reposted by Ben Golub

Harvey Lederman @harveylederman.bsky.social · Aug 5

I wrote about automation and the meaning of life, as a guest post on Scott Aaronson's Shtetl-Optimized. (1/5)

scottaaronson.blog?p=9030

ChatGPT and the Meaning of Life: Guest Post by Harvey Lederman

Scott Aaronson’s Brief Foreword: Harvey Lederman is a distinguished analytic philosopher who moved from Princeton to UT Austin a few years ago. Since his arrival, he’s become one of my …

scottaaronson.blog

1 5 48

Reposted by Ben Golub

Charlie Cheever @ccheever.com · Aug 4

A 16 y o running a 1:42.27 800m is maybe the most impressive and amazing and shocking athletic feat I have ever seen. Hard to explain how truly extraordinary it is.

Travis Miller @travismillerx13.bsky.social · Aug 3

😤DONAVAN BRAZIER 1:42.16 PB TO WIN 800M NATIONAL TITLE!!!!

1:42.27 U18 WORLD RECORD FOR 16-YEAR-OLD COOPER LUTKENHAUS IN SECOND!!!

1 11

Reposted by Ben Golub

Juan Diego Rodriguez (@ COLM 2025) @juand-r.bsky.social · Aug 1

Reminds me of this paper aclanthology.org/2025.finding...

Large Language Models Reflect Human Citation Patterns with a Heightened Citation Bias

Andres Algaba, Carmen Mazijn, Vincent Holst, Floriano Tori, Sylvia Wenmackers, Vincent Ginis. Findings of the Association for Computational Linguistics: NAACL 2025. 2025.

aclanthology.org

2 5

Ben Golub @bengolub.bsky.social · Aug 1

On the one hand, LLM's are likeliest to know about papers that have been summarized well tens or hundreds of times, and suggest them appropriately.

On the other hand, their use crowds out asking a colleague, and the colleague might be even more likely to name a seminal/focal paper.

2 6

Ben Golub @bengolub.bsky.social · Aug 1

I wonder how LLM literature reviews will change "rich-get-richer" effects in academic citations.

2 18

Reposted by Ben Golub

Eva Vivalt @evavivalt.bsky.social · Aug 1

Okay, now that I've used it for some work on forecasting, I have to say this is a fantastic tool! We will be changing some things in our paper as a result. Highly recommended!

Ben Golub @bengolub.bsky.social · Jul 24

I've been working on a new tool, Refine, to make scholars more productive. If you're interested in being among the very first to try the beta, please read on.

Refine leverages the best current AI models to draw your attention to potential errors and clarity issues in research paper drafts.

1/

1 1 11

Ben Golub @bengolub.bsky.social · Aug 1

Wow - thanks, Eva!!

Reposted by Ben Golub

Marciano Siniscalchi @marciano.bsky.social · Jul 30

Belatedly doing my best to amplify this. Refine works, and I can totally see it becoming part of my pre-submission routine.

Ben Golub @bengolub.bsky.social · Jul 24

I've been working on a new tool, Refine, to make scholars more productive. If you're interested in being among the very first to try the beta, please read on.

Refine leverages the best current AI models to draw your attention to potential errors and clarity issues in research paper drafts.

1/

2 6

Ben Golub @bengolub.bsky.social · Jul 30

YAY :)

Reposted by Ben Golub

Jessica Leight @leightjessica.bsky.social · Jul 30

Refine is at a completely different level from the other LLM-based review tools I have tried. It not only identified a large number of small errors in my paper, it also correctly identified what I think is the biggest conceptual weakness. Definitely worth trying.

Ben Golub @bengolub.bsky.social · Jul 24

I've been working on a new tool, Refine, to make scholars more productive. If you're interested in being among the very first to try the beta, please read on.

Refine leverages the best current AI models to draw your attention to potential errors and clarity issues in research paper drafts.

1/

1 4 17

Reposted by Ben Golub

Ryan Briggs @ryancbriggs.net · Jul 29

fwiw I used refine & while it misread my paper & made mistakes in 2 places, they were pretty "good" mistakes & I'm going to change the text to be clearer because I think a human could make the same mistake. The other comments were also helpful. It's the best "LLM peer review" experience I've found.

1 6

Reposted by Ben Golub

Max Reith @maxreith.bsky.social · Jul 28

Pretty impressive! It caught things that o3 and Gemini didn't. Wish I had had this before my submission...

Ben Golub @bengolub.bsky.social · Jul 24

I've been working on a new tool, Refine, to make scholars more productive. If you're interested in being among the very first to try the beta, please read on.

Refine leverages the best current AI models to draw your attention to potential errors and clarity issues in research paper drafts.

1/

1 6

Reposted by Ben Golub

Economists for Ukraine @econ4ua.bsky.social · Jul 25

👩‍🏫The war in Ukraine has devastated the education system, so the AI for Good Foundation and Econ4UA created education centers in four frontline cities. Located in air raid shelters with reliable Internet, they offer free classes and a safe space for children to learn.

Support our centers:

Donate: Support Ukraine's Education - AI for Good Foundation

Ukrainian Education: Supporting Ukraine’s Children in Conflict Zones—From Safe Learning Spaces to Essential Life Skills. Ongoing Campaign. Learn More

ai4good.org

2 6

Reposted by Ben Golub

Seema Jayachandran @seema.bsky.social · Jul 25

I just tried out Refine on a paper of mine and found the comments very helpful.

Ben Golub @bengolub.bsky.social · Jul 24

I've been working on a new tool, Refine, to make scholars more productive. If you're interested in being among the very first to try the beta, please read on.

Refine leverages the best current AI models to draw your attention to potential errors and clarity issues in research paper drafts.

1/

3 1 19