Luigi Mangione Being Charged as a Terrorist Is Borderline Prosecutorial Misconduct
Added 2024-12-20 06:16:21 +0000 UTCOA1102 - Why is the state of New York charging Luigi Mangione as a terrorist? And why have federal charges been filed in this case at all? We consider before moving on to try to locate any actual legal basis in the Georgia Court of Appeals decision removing Fulton County DA Fani Willis from the prosecution of Donald Trump and his alleged conspirators in Georgia. Finally, we dig into 11th Circuit judge Kevin Newsom’s surprisingly defensible argument that judges and lawyers should occasionally consult with ChatGPT, and Matt drops a footnote involving judges being weird about food.
New York indictment of Luigi Mangione (12/18/24)
Federal indictment of Luigi Mangione (12/19/24)
Georgia Court of Appeals decision removing Fani Willis from Trump’s GA case (12/19/24)
Snell v. Specialty Insurance Corp, 11th Cir. (5/28/2024)(Newsom, J., concurring)
U.S. v. Deleon, 11 Cir. (9/5/2024)(Newsom, J., concurring)
Judge Proctor’s order in McCuller v. Koch (11/26/24)
Ohio Supreme Court’s decision in Berkheimer v. REKM (7/25/24)
Ohio Supreme Court’s decision denying reconsideration in Berkheimer v. REM (12/9/24)
Check out the OA Linktree for all the places to go and things to do!
Comments
I was there too! My whole office (ok, there are only 3 of us) went. Small world!
Katie Herrmann
2024-12-30 01:36:17 +0000 UTCThat’s awesome! I was at the Hamilton for 5-4’s DC live show. I’ve been here on Capitol Hill for coming on 7 years now. A meet up sounds like a treat. Not to sound creepy, but I’ll find you a more social media 👍.
Austin Flake
2024-12-29 23:48:31 +0000 UTCHi Austin! Good to see a fellow 5-4 fan. One of our Associate attorneys is pretty active on the 5-4 Slack (that's actually where we met her). I'm one of the "DC freaks" - which is a solid group of folks that meets up from time to time. Happy to welcome you in if you're from the area!
Katie Herrmann
2024-12-29 01:06:20 +0000 UTCI listened to this episode with my husband, a post-doctoral researcher who builds LLMs which analyse sentiment and opinions in writing. He recommends that you speak to a researcher in the area of LLMs+law area regarding this op-ed. My husband is no authority of course, but he said there are problems in the judge's opinion which are known to people who research LLMs+law. I'm personally extremely interested now in this further and whilst I will be spending the Christmas break picking his brains on this and looking into the literature myself, I'd like even more an OA ep on it hehe. I told my husband he should speak to you about this, but unfortunately he works for a government org and is forbidden from commenting until his contract is up :') huzzah. Anyways great episode as always!
Jess C
2024-12-27 02:34:20 +0000 UTCThanks for letting me know! just pitched them on a couple of show ideas. Really do appreciate this
Matt Cameron
2024-12-23 04:01:06 +0000 UTCMatt, I don’t know if you listen to the 5-4 podcast. They just said on their Patreon that they’d really like to have an immigration attorney on to interview.
Austin Flake
2024-12-23 00:42:12 +0000 UTCI have a t-shirt that says "Google Jury Nullification," which I only mention now because I am crazy random.
Katie Herrmann
2024-12-21 04:28:36 +0000 UTCI'd treat it like a dictionary in that if the A.I. is the best argument/evidence I've got to support my side, I'll try to settle. But if I have otherwise strong arguments, I might tack it on to the laundry list as kind of a "even ChatGPT when asked agrees that . . ." That seems like it's where Newsom is at, which seems harmless if not particularly beneficial.
Gmork
2024-12-20 22:45:58 +0000 UTCThanks for putting this into words. I’m pretty empathetic to most people but feel pretty much nothing in this case.
Katie Byrum
2024-12-20 22:33:55 +0000 UTCthis is an extremely helpful perspective! From my (and, I think, Judge Newsom's) non-technical viewpoint ChatGPT seems like it represents an entire world, but it's good to remember that it is really just an extremely sophisticated guesser. I still think Newsom has some good points in those concurrences (and I'd be interested to know what you think if you want to take a look at the way he lays it all out) but it also seems like he really consult with anyone who actually understands what's behind the curtain.
Matt Cameron
2024-12-20 21:50:04 +0000 UTCIf anyone is discriminating against folks from a protected class, its insurance companies. If you go to the ER with severe abdominal pain, you better pray its your appendix and not a serious ovarian cyst. They'll cover treatment of one before it ruptures and becomes life-threatening. Guess which one. Gender affirming care? Thats going to be a prolonged battle. I worked at a restaurant more than a decade ago with a guy who bought cheap insurance bc its what he could afford. His wife became pregnant and he learned only then that they didnt cover pregnany. What?!? He'd been paying them money he couldnt afford every month for years. Insane.
Katie Herrmann
2024-12-20 21:50:01 +0000 UTCAs an insurance lawyer, I think you're usually on the losing side when you find yourself citing the dictionary. Context, pattern and practice, and pure intuition are all much more persuasive ways of arguing what the intent of contract language was. When you find yourself breaking open a dictionary for non-technical words, it's usually bc those things don't support you.
Gmork
2024-12-20 17:56:42 +0000 UTCI was really surprised how many lawyers were celebrating that Go To Lunch order. If anyone should appreciate the inherent problems in a judge getting creative and ordering the lawyers before him to dance, it should be those lawyers. Also seemed weird bc it didn't seem like the defense attorney in that case did anything wrong.
Gmork
2024-12-20 16:57:25 +0000 UTCA comment about large language models and dictionaries. Present-day dictionaries are not "one man's" opinion. Lexicographers use a corpus of representative texts to determine how a word or phrase is used (and yes, dictionaries do include phrases when they become usual formulations). Black's Law Dictionary (12th ed.), for example, defines "physical restraint" as "a hard-to-remove object or device that is placed on a person to restrict the person's freedom of movement or normal access to the person's own body." This definition would presumably be derived from how the phrase is used in judicial opinions, law review articles, and other legal texts. Different dictionaries will use different corpora depending on their editorial requirements. Merriam-Webster, for instance, relies primarily on American texts for the general public that have been published and edited; the Oxford English Dictionary covers global English with a bias toward British English. Hence different dictionaries will have slightly different definitions, and one should chose the one most appropriate for one’s purpose. Note that this does not necessarily apply to older dictionaries that may be used to determine common usage in the when the founders wrote the constitution. Samuel Johnson's dictionary (1755) and Noah Webster's (1828) were indeed basically one-man shows using a limited corpus of material, if at all. For example, linguists have conducted extensive reviews of 18th century material regarding the phrase "bear arms" and have determined that Scalia was absolutely wrong in Heller; the term as commonly used then referred almost exclusively to military service. And note that the term "Webster” is essentially meaningless. The term is not trademarked and a number of dictionaries from different publishers use that name. Merriam-Webster is the most famous and the present-day descendent of Noah Webster’s original. Large-language models, like ChatGPT, also rely upon a corpus, but there is little or no selection of texts to ensure they are representative; they are simply scraped en masse from the internet and publisher databases. (More is not not necessarily better. One must take into account where the texts come from, what their nature is, and when they were written to create a valid corpus.) Furthermore, the analysis is very different than a lexicographer's. LLM-based AIs simply determine what terms are most likely to appear next in a sequence. There is no understanding of meaning or context. Furthermore, the entire process is a black box, and even the creators cannot explain how the AI produces its results, unlike a human lexicographer who can explain their sources and reasoning. There is one other major problem with LLM-based AI that is just starting to appear. As more and more texts are being produced by LLM-based AIs, those texts are “contaminating” the training data used by newer generations of LLM-based AIs. ChatGPT is starting to tell users how AIs use language, not how actual humans use and understand it. It’s a positive feedback loop, which as any engineer will tell you will ultimately result in undesirable, if not outright disastrous, results. LLM-based AIs are simply not suited to this type of work, which is the problem generally with their mass deployment. They are exquisitely well-suited for one particular purpose (i.e., guessing what word most likely would come next), and are terrible at most of the purposes for which people are trying to use them.
Wordorigins.org
2024-12-20 16:54:29 +0000 UTCDid the Georgia appellate court bother with any theory about how Wade and Willis' relationship prejudiced Trump? Are they going with, "she prosecuted Trump only so that she coukd hire her bf as special counsel" ?
Gmork
2024-12-20 15:59:08 +0000 UTCVoir dire is going to be wild trying to exclude people with no bad experiences or opinions about healthcare.
Gmork
2024-12-20 15:57:23 +0000 UTCCould be as simple as someone on the jury not getting a promotion with the enhancement. We also don't know how their inevitable interactions with insurance can color them.
Wishing every Clarence Thomas a very God rest his soul
2024-12-20 15:55:03 +0000 UTCI'm not optimistic that he'll get a hung jury, but if he does, it'll be in large part bc prosecutors overreach in trying to appease their donors w a big show rather than just trying a basic murder case.
Gmork
2024-12-20 15:32:45 +0000 UTCYes it seems like they are waaay over-charged. Esp considering what those who knew him say about his mental state lately
Reese
2024-12-20 15:21:31 +0000 UTCThe 'victim' was a man who made a conscious choice to kill his customers for money - and somehow that isn't murder in our weird money worshipping world. Apparently it's fine if you use a spread sheet. It's not just in the US that there is anger either. We've all spent 60 years watching these 'people' asset strip the world while we get poorer and we're supposed to thank them for their effing service? Pretend to be sad when they get a little karma? Thoughts and prayers? Do I have to pretend that if Murdoch or Musk or the remaining Koch brother were next I wouldn't raise a glass? It's weird because I don't think I knew I'd reached this level of disaffection until this happened. I'm pretty against murder in most circumstances but I find myself asking whether justifiable homicide is a thing that could apply in situations like this. Morally if not legally.
Tom Pegg
2024-12-20 13:24:48 +0000 UTC
