Greetings, my name is David Colarusso. I'm the co-director of Suffolk University Law School's Legal Innovation & Technology (LIT) Lab. With one foot in law and the other in tech, I really want the open web to thrive. So I created a bot (@icymilaw.org) and this site to help folks discover great law-themed content while showing off what one can do with sufficiently open protocols.
If you like these, you'll ❤️ The Finite Scroll, an open-source client-side algorithmically-driven RSS reader. You might also enjoy this post: How and why I (still) use social media. It includes tips on how to make your own custom social media algo(s).
Note, the number of fire emoji represent how many standard deviations more popular a link is than the average link observed in its category.
News-like Links
A collection of links shared recently¹ by legal-type folks² with URLs that look like they point to news articles,³ sorted by popularity.
- Appeals court clears way for U.S. to reopen border for asylum seekers 🔥🔥🔥🔥🔥🔥A federal appeals court found Trump’s declaration of an “invasion” at the U.S.-Mexico border was illegal, effectively reopening the U.S. to migrants seeking asylum.
- Justice Department Drops Criminal Investigation into Fed 🔥🔥🔥
- Elon Musk’s near-daily online posts about race are turning off some fans 🔥Elon Musk posted almost daily about race on X in recent months, a Washington Post analysis found, at times sharing views experts said echoed white supremacy.
- Britain and Spain Reject Reported Plans by Trump to Punish Them 🔥
- US millionaire big-game hunter dies after being crushed by elephants 🔥Ernie Dosio, a 75-year-old vineyard owner, was hunting an antelope species in Africa when the incident occured ...
- Shabana Mahmood does not rule out sending back refused Afghan asylum seekers 🔥Home secretary indicates Whitehall talks about returns programme, a move that would shock humanitarian groups ...
- Maine Governor Mills Vetoes Statewide Data Center Moratorium 🔥Maine Governor Janet Mills vetoed what would have been the first statewide freeze on large data center development, saying it would hurt a part of Maine in need of an economic boost.
- White House Allowed Officials’ Text Messages to Be Deleted, Lawsuit Says 🔥
- Mandelson under formal investigation by EU's anti-fraud office 🔥The EU's anti-fraud office confirms it has ...
- Exclusive: Pentagon email floats suspending Spain from NATO, other steps over Iran rift, source says 🔥The internal email outlines options for the United States to punish NATO allies it believes failed to support U.S. operations in the war with Iran, including reviewing the U.S. position on Britain's ...
Blog-like Links
A collection of links shared recently⁴ by legal-type folks⁵ with URLs that look like they point to blogs/newsletters,⁶ sorted by popularity.
- What’s Wrong With The SPLC Indictment 🔥🔥🔥DOJ's indictment of the Southern Poverty Law Center is a critical moment in the administration's war on democracy ...
- Uxoricide: The Word You Don't Know Because You're Not Supposed To 🔥On the Language We Don't Have for Men Who Kill Their Wives.
- Trump Cuts the Inspectors General 🔥The Trump administration’s proposal to drastically cut the funding of the inspectors general system would damage one of the federal government’s most effective budgeting mechanisms.
- Texas Just Mandated Protestant Christianity in Every Public School 🔥The 5th Circuit upheld SB 10 by one vote. The majority called a 16-by-20-inch Protestant religious poster on every public school wall a “passive display.”
- NEWS: White House Braces for 'Political Collapse,' DOJ Brings Back Modern-Day Gas Chamber, Patel Previously Arrested for Intoxication Good evening, everyone.
AI & The Law Links
A collection of links shared recently⁷ on Bluesky that look like they talk about AI & the law,⁸ sorted by popularity.
- The Delusion of ‘AI Justice’† 🔥🔥🔥🔥As AI seeps deeper into our judicial system, boosters insist it will bring both fairness and efficiency. But can we really trust Judge Grok?
- Florida police used Flock cameras to track No Kings protesters 🔥🔥Florida law enforcement agencies have used Flock cameras to track drivers tied to political protests and rallies throughout the state, police records show.
- Elite law firm Sullivan & Cromwell admits to AI ‘hallucinations’ 🔥Firm whose partners bill more than $2,000 per hour apologises to judge for software-driven errors in bankruptcy case ...
- AI Has Emboldened Child Predators, and Investigators Can't Keep Up† 🔥To find real kids in danger, law enforcement must sift through a surge of AI-generated sex imagery.
- Altman apologizes after OpenAI failed to alert police before Tumbler Ridge killings 🔥OpenAI's head, Sam Altman, has apologized for not alerting law enforcement about the online behavior of a person who killed eight people in Tumbler Ridge, British Columbia.
Law Review-like Links
A collection of links shared recently⁹ on Bluesky that look like they point to papers in law journals or the like,¹⁰ sorted by popularity.
- Everything-Everywhere Searches and the Geofence Puzzle† 🔥🔥Police surveillance technologies have outpaced Fourth Amendment doctrine. Arbitrary, generalized search powers stretching across entire cities and states now ex ...
- A Foundation, Not a Finish Line: Statewide Structure and the Future of Public Defense This keynote address is, at its core, a love letter to public defense — to the defenders who stand beside accused people every day, force the system to recog ...
- A Systematic Content Analysis of Innovation in European Competition Law Innovation plays a crucial role in defining competitive dynamics. Given this fact, one might expect ‘innovation’ to play a consistent role in antitrust law. The ...
AI Paper-like Links
A collection of links shared recently¹¹ on Bluesky that look like they point to papers on AI,¹² sorted by popularity. Wondering why this section is on a site about the law? Well, I teach a course on AI & the Law, and it turns out that understanding this stuff is super important to figuring out what the law might have to say. So, I figured since I was sharing lists, I might as well share this one too.
- LLMs Corrupt Your Documents When You Delegate† 🔥🔥🔥Large Language Models (LLMs) are poised to disrupt knowledge work, with the emergence of delegated work as a new interaction paradigm (e.g., vibe coding). Delegation requires trust - the expectation t...
- Critical Confabulation: Can LLMs Hallucinate for Social Good?† 🔥LLMs hallucinate, yet some confabulations can have social affordances if carefully bounded. We propose critical confabulation (inspired by critical fabulation from literary and social theory), the use...
- Ask don't tell: Reducing sycophancy in large language models† 🔥Sycophancy, the tendency of large language models to favour user-affirming responses over critical engagement, has been identified as an alignment failure, particularly in high-stakes advisory and soc...
- There Will Be a Scientific Theory of Deep Learning In this paper, we make the case that a scientific theory of deep learning is emerging. By this we mean a theory which characterizes important properties and statistics of the training process, hidden representations, final weights, and performance of neural networks. We pull together major strands of ongoing research in deep learning theory and identify five growing bodies of work that point toward such a theory: (a) solvable idealized settings that provide intuition for learning dynamics in realistic systems; (b) tractable limits that reveal insights into fundamental learning phenomena; (c) simple mathematical laws that capture important macroscopic observables; (d) theories of hyperparameters that disentangle them from the rest of the training process, leaving simpler systems behind; and (e) universal behaviors shared across systems and settings which clarify which phenomena call for explanation. Taken together, these bodies of work share certain broad traits: they are concerned with the dynamics of the training process; they primarily seek to describe coarse aggregate statistics; and they emphasize falsifiable quantitative predictions. We argue that the emerging theory is best thought of as a mechanics of the learning process, and suggest the name learning mechanics. We discuss the relationship between this mechanics perspective and other approaches for building a theory of deep learning, including the statistical and information-theoretic perspectives. In particular, we anticipate a symbiotic relationship between learning mechanics and mechanistic interpretability. We also review and address common arguments that fundamental theory will not be possible or is not important. We conclude with a portrait of important open directions in learning mechanics and advice for beginners. We host further introductory materials, perspectives, and open questions at learningmechanics.pub.
- Closing the Domain Gap in Biomedical Imaging by In-Context Control Samples The central problem in biomedical imaging are batch effects: systematic technical variations unrelated to the biological signal of interest. These batch effects critically undermine experimental reproducibility and are the primary cause of failure of deep learning systems on new experimental batches, preventing their practical use in the real world. Despite years of research, no method has succeeded in closing this performance gap for deep learning models. We propose Control-Stabilized Adaptive Risk Minimization via Batch Normalization (CS-ARM-BN), a meta-learning adaptation method that exploits negative control samples. Such unperturbed reference images are present in every experimental batch by design and serve as stable context for adaptation. We validate our novel method on Mechanism-of-Action (MoA) classification, a crucial task for drug discovery, on the large-scale JUMP-CP dataset. The accuracy of standard ResNets drops from 0.939 $\pm$ 0.005, on the training domain, to 0.862 $\pm$ 0.060 on data from new experimental batches. Foundation models, even after Typical Variation Normalization, fail to close this gap. We are the first to show that meta-learning approaches close the domain gap by achieving 0.935 $\pm$ 0.018. If the new experimental batches exhibit strong domain shifts, such as being generated in a different lab, meta-learning approaches can be stabilized with control samples, which are always available in biomedical experiments. Our work shows that batch effects in bioimaging data can be effectively neutralized through principled in-context adaptation, which also makes them practically usable and efficient.
The High Score
The 20 accounts most reposted by @icymilaw.org over the past week¹³ (the list below is updated every Sunday). High Score, get it? One Score = 20, as in, "four score and seven years ago." ;)
- Law + Tech News Bot (@news.bot.suffolklitlab.org)
- ICYMI (Law) (@icymilaw.org)
- Joyce White Vance (@joycewhitevance.bsky.social) (promoted)
- New York Attorney General Letitia James (@newyorkstateag.bsky.social) (promoted)
- National Security Counselors 🕵 (@nationalsecuritylaw.org) (promoted)
- Raffi Melkonian (@rmfifthcircuit.bsky.social) (promoted)
- Evan Bernick, a finite mode with a smol hooman and a lorg floof (@evanbernick.bsky.social)
- Alejandra Caraballo (@esqueer.net) (promoted)
- Quinta Jurecic (@qjurecic.bsky.social) (promoted)
- ElieNYC (@elienyc.bsky.social) (promoted)
- Aaron Reichlin-Melnick (@reichlinmelnick.bsky.social) (promoted)
- John Pfaff (@johnpfaff.bsky.social) (promoted)
- Daniel Suitor (@danielsuitor.com) (promoted)
- Matt Ford (@mford.bsky.social) (promoted)
- Just Security (@justsecurity.org) (promoted)
- Professa Murray (@kalimurray.bsky.social) (promoted)
- dag (@davidallengreen.bsky.social) (promoted)
- Joe Patrice (@joepatrice.bsky.social) (promoted)
- Daniel Gilmore (@gilmored85.bsky.social) (promoted)
- Melissa Gira Grant (@melissagiragrant.com)
- Walter Olson (@walterolson.bsky.social) (relegated)
- AG Andrea Joy Campbell (@massago.bsky.social) (relegated)
- Don Moynihan (@donmoyn.bsky.social) (relegated)
- Michael J. Stern (@michaeljstern.bsky.social) (relegated)
- Kevin M. Kruse (@kevinmkruse.bsky.social) (relegated)
- No Further Force Or Popehat (@kenwhite.bsky.social) (relegated)
- Katie Phang (@katiephang.bsky.social) (relegated)
- Bill Grueskin (@bgrueskin.bsky.social) (relegated)
- Anthony Michael Kreis (@anthonymkreis.bsky.social) (relegated)
- Ryan Goodman (@rgoodlaw.bsky.social) (relegated)
- Sarah Fackrell (@design-law.bsky.social) (relegated)
- Ray Beckerman (@raybeckerman.bsky.social) (relegated)
- Cristian Farias (@cristianfarias.com) (relegated)
- Corey Rayburn Yung (@coreyryung.bsky.social) (relegated)
- Gabriel Malor (@gabrielmalor.bsky.social) (relegated)
- Marty Lederman (@martylederman.bsky.social) (relegated)
† This link was also in yesterday's digest. ¹ Approx. 1 day lookback. ² Attorneys, law profs, et al. ³ News-like links (law) ⁴ Supra note 1. ⁵ Supra note 2. ⁶ Blog-like links (law) ⁷ Approx. 3.5 days lookback. ⁸ AI & the Law ⁹ Approx. 1 week lookback. ¹⁰ Law Review-like ¹¹ Supra note 9. ¹² AI Papers et al. ¹³ High Score
Search
Use Google's site search to look through prior digests.