SWEET REVENGE > "Instead, if the actions Trump took irritated Swift, the sweetest revenge would be a highly publicized endorsement of Harris
Taylor Swift endorses Harris in post signed 'Childless Cat Lady'
Taylor Swift Responded To Trump’s AI Driven Nonsense With An Endorsement Rather Than A Lawsuit
from the as-we-were-saying dept
Last month, we discussed the internet’s reaction to Donald Trump, well, Donald Trumping all over social media. He shared several images on social media, some of which were real, some of which were parody, and some of which were AI generated images, all of which appeared to suggest that Taylor Swift had endorsed him. In fact, his own contribution to that social media post were two whole words: “I accept.”
But there was nothing to accept, of course. And the internet’s reaction in far too many places was all in advocating legal action by Swift’s team, including utilizing new laws untested in the courtroom in order to sue Trump over this implied false endorsement. I suggested that would be a massive waste of time and money. Instead, if the actions Trump took irritated Swift, the sweetest revenge would be a highly publicized endorsement of Harris (Swift endorsed Biden last cycle) or, in lieu of that, at least a public rebuke of the social media posts. From that previous post:
But the larger point here is that all Swift really has to do here is respond, if she chooses, with her own political endorsement or thoughts. It’s not as though she didn’t do so in the last election cycle. If she’s annoyed at what Trump did and wants to punish him, she can solve that with more speech: her own. Hell, there aren’t a ton of people out there who can command an audience that rivals Donald Trump’s… but she almost certainly can!
Just point out that what he shared was fake. Mention, if she wishes, that she voted against him last time. If she likes, she might want to endorse a different candidate.
Now, I’m quite certain Her Swiftness didn’t actually read my post and take it as the advice it was designed to be, but she certainly has acted as though she had. Less than a month after Trump’s AI nonsense, and immediately after the debate between Trump and Harris, Taylor Swift took to social media herself both to address what Trump did and to endorse Harris.
Recently I was made aware that AI of ‘me’ falsely endorsing Donald Trump’s presidential run was posted to his site. It really conjured up my fears around AI, and the dangers of spreading misinformation. It brought me to the conclusion that I need to be very transparent about my actual plans for this election as a voter. The simplest way to combat misinformation is with the truth.
I will be casting my vote for Kamala Harris and Tim Walz in the 2024 Presidential Election. I’m voting for @kamalaharris because she fights for the rights and causes I believe need a warrior to champion them. I think she is a steady-handed, gifted leader and I believe we can accomplish so much more in this country if we are led by calm and not chaos. I was so heartened and impressed by her selection of running mate @timwalz, who has been standing up for LGBTQ+ rights, IVF, and a woman’s right to her own body for decades.
I’ve done my research, and I’ve made my choice. Your research is all yours to do, and the choice is yours to make. I also want to say, especially to first time voters: Remember that in order to vote, you have to be registered! I also find it’s much easier to vote early. I’ll link where to register and find early voting dates and info in my story.
With love and hope,
Taylor Swift
Childless Cat Lady
Now, the focus of this post is not that Trump is the Big Bad, or to suggest that anyone view Swift’s endorsement as a gospel they themselves must follow. Instead, the point is the one we made originally: no lawsuit was needed for Swift to address this. As we said: “she can solve [this] with more speech: her own.” That lines up fairly well with Swift’s statement: “The simplest way to combat misinformation is with the truth.”
The real question is whether this endorsement ever would have happened without Trump’s online antics. We can’t know for certain, but it sure sounds from Swift like they may have lit the spark.
Filed Under: ai, deepfakes, donald trump, endorsement, generative ai, kamala harris, more speech, response, taylor swift
Court Tosses Lawsuit From Woman Wrongly Jailed For 2 Weeks Because ‘Bivens’ Precedent Only Works For Webster Bivens
from the what-was-even-the-point-of-the-original-decision-then dept
If you think qualified immunity is awarded far too frequently during far too many cases involving obvious rights violations, brace yourself for the harrowing insanity that is the wreckage that has been made of Supreme Court’s Bivens decision.
That case involved rights violations by federal officers against Webster Bivens. The Supreme Court found in favor of Bivens and created a cause of action that allowed people to sue federal officers in court. It shouldn’t have bothered. Since then, almost zero federal courts have allowed plaintiffs to bring lawsuits under this precedent. Any time one of the lower courts does, the Supreme Court reminds them that a Bivens lawsuit’s facts have to be almost exactly the same as those involving the original plaintiff in the 1971 case. Anything else is considered to be creating a “new” cause of action and strictly forbidden.
And so it goes here in the case of Judith Henry, who was arrested by US Marshals and held for two weeks (in two different jails) before someone finally decided to run her prints. Once that happened, law enforcement finally realized they had the wrong person in custody.
The Third Circuit — like pretty much every other federal court — says Henry can’t sue the Marshals who arrested her. The facts of the case, though, are absolutely flabbergasting. From the decision [PDF]:
In 1993, a different woman named Judith Maureen Henry (the “Absconder”) skipped parole in Pennsylvania.
Fast forward to 2019, when the director of the Pennsylvania Interstate Parole Services issued a warrant for the Absconder’s arrest for the parole violation. That warrant, however, targeted Henry’s home address and attached her driver’s license photo. Parole Services forwarded the warrant to officials in New Jersey.
It’s not the most egregious case of mistaken identity. Both the parole violator and the innocent woman had the same name. But it does seem extremely absurd that Pennsylvania parole services is issuing warrants for someone it apparently hadn’t checked up on in more than two decades. Given that massive time gap, more due diligence was needed to ensure the wrong person didn’t get picked up, which obviously didn’t happen here.
Instead, the information passed on to the Marshals Service told the officers to find this woman at this address. Which they did. She protested during her arrest, but there was no reason (at that time) for the involved officers to doubt they had the correct person.
But they could have followed up rather quickly and painlessly on her protests. She repeatedly asked them to take her prints and compare them to those of the absconded Judith Henry. These requests were ignored until she was transferred back to Pennsylvania. The wrong Judith Henry was arrested on August 22. Her prints weren’t taken until September 3. Then, despite knowing her prints weren’t a match, Pennsylvania law enforcement kept her in jail until September 5.
The appeals court says the Marshals were justified in relying on the arrest warrant, especially since the arrested person had the same name as the parole violator they were looking for. But even if these violations — including the extended detention — were firmly linked to the Marshals (who only effected the arrest), this wouldn’t have worked out for the innocent Judith Henry. And that finding still holds even if the Third Circuit agrees (which it does) that the state officers working with the Marshals can be treated as federal officers for the purposes of this lawsuit.
The precedent is so narrow it only applies to one person: and that person is the one who helped create the precedent that apparently is now impossible to apply to anyone else for any reason.
The only Supreme Court case authorizing a Bivens remedy for police misconduct is Bivens itself. Ziglar, 582 U.S at 131.3 The complaint in Bivens alleged that the plaintiff was invalidly subject to a warrantless arrest in his home. 403 U.S. at 389. We do not read Henry’s complaint as sufficiently alleging that state of affairs.
As the decision notes, the Supreme Court has only authorized a Bivens action three times since its 1971 ruling. Its direction to lower courts is to reject any “extension” or “new context” if judges see “any reason to pause” before attempting to expand this precedent. How that works out for plaintiffs is that the government always wins, no matter what violations federal officers have committed. There’s just not enough space left in the jurisprudence for plaintiffs to get a cause of action in edgewise.
A context is new if it “is different in a meaningful way from previous Bivens cases decided by” the Supreme Court. Ziglar v. Abbasi, 582 U.S. 120, 139 (2017). Whether a context is new is an “easily satisfied” test because “a modest extension [of the Bivens action] is still an extension.” Id. at 147-49. Even “significant parallels to one of the [Supreme] Court’s previous Bivens cases” may not be enough.
That shuts down pretty much the entire lawsuit. As for Henry’s far-from-unreasonable requests she be fingerprinted to clear up the confusion, the court has this to say about this request that went ignored for most of two weeks:
We grant that, asking those questions in this case, a reasonable observer could conclude the answers are not hard to find and would impose minimal burdens on the Marshals. Henry’s request was modest: merely that her fingerprints be compared to the Absconder’s. But it is for Congress, not the judiciary to “balance[] the costs and benefits” of a cause of action against the Marshals bottomed on their failure to investigate Henry’s claims of innocence. The legislature must decide whether the “potential encroachment” on the executive branch’s investigatory function “is worth it.” Accordingly, concerns for separation of powers counsel hesitation before we endorse a Bivens suit in a new context.
And there it is: it’s up to Congress to fix this. Well, we won’t be holding our breath. If there’s anything Congress isn’t in any hurry to do, it’s making it easier to sue federal employees for rights violations. This decision is just another nail in the Bivens coffin. It will only take a few more before the precedent ceases to exist entirely.
Filed Under: 14th amendment, 3rd circuit, 4th amendment, bivens, false arrest
More Of RFK Jr.’s ‘Don’t Moderate Me, Bro’ Cases Are Laughed Out Of Court
from the that's-not-how-any-of-this-works dept
In the last month, I wrote about two of Robert F. Kennedy Jr.’s bullshit crazy lawsuits over him being very, very mad that social media companies keep moderating or limiting the spread of his dangerous bullshit anti-vax nonsense. In one, the Ninth Circuit had to explain (not for the first time) to RFK and his disgraced Yale Law professor lawyer, Jed Rubenfeld, that Meta fact checking RFK Jr. does not violate the First Amendment, and that Section 230 does not turn every internet company into a state actor.
In the other case, one of the MAGA world’s favorite judges ignored both the facts and the scolding he just got from the Supreme Court to insist that the Biden administration has been trying to censor RFK Jr., a thing that has not actually happened.
But Professor Eric Goldman reminds me that there were two other cases involving RFK Jr. and his anger at being moderated that had developments that I hadn’t covered. And both of them were, thankfully, not in courtrooms of partisan judges who live in fantasylands.
First, we had a case in which RFK Jr. sued Meta again. I had mentioned this case when it was filed. The Ninth Circuit one mentioned above was also against Meta, but RFK Jr. decided to try yet again. In this case, he also sued them claiming that efforts to restrict a documentary about him by Meta violated his First Amendment rights.
If you don’t recall, Meta very temporarily blocked the ability to share the documentary, which they chalked up to a glitch. They fixed it very quickly. But RFK Jr. insisted it was a deliberate attempt to silence him, citing Meta’s AI chatbot as giving them the smoking gun (yes, they really did this, even the chatbot is just a stochastic parrot spewing whatever it thinks will answer a question).
What I had missed was that district court Judge William Orrick, who is not known for suffering fools lightly, has rejected RFK Jr.’s demands for a preliminary injunction. Judge Orrick is, shall we say, less than impressed by RFK Jr. returning to the well for another attempt at this specious argument, citing the very Ninth Circuit case that RFK Jr. just lost in his other case against Meta.
The plaintiffs assert that they are likely to succeed on the merits of their First Amendment claim, which is that Meta violated their rights to free speech by censoring their posts and accounts on Meta’s platforms. But the First Amendment “‘prohibits only governmental abridgment of speech’ and ‘does not prohibit private abridgment of speech.’” Children’s Health Def. v. Meta Platforms, Inc., —F. 4th—, No. 21-16210, 2024 WL 3734422, at *4 (9th Cir. Aug. 9, 2024) (first quoting Manhattan Cmty. Access Corp. v. Halleck, 587 U.S. 802, 808 (2019); and then citing Prager Univ. v. Google LLC, 951 F.3d 991, 996 (9th Cir. 2020)). Because there is no apparent state action, this claim is unlikely to succeed.
RFK Jr. twists himself into a pretzel shape to try to claim that Meta is magically a state actor, but the court has to remind him that these arguments are quite stupid.
The Ninth Circuit recently has twice affirmed dismissal of claims filed by plaintiffs alleging that social media platforms violated the plaintiffs’ First Amendment rights by flagging, removing, or otherwise “censoring” the plaintiffs’ content shared on those platforms. See Children’s Health, 2024 WL 3734422 at *2–4; O’Handley, 62 F.4th at 1153–55. In both cases, the Ninth Circuit held that the plaintiffs’ claims failed at the first step of the state action framework because of “the simple fact” that the defendants “acted in accordance with [their] own content-moderation policy,” not with any government policy…..
The only difference between those cases and this one is that here, the plaintiffs seem to allege that the “specific” harmful conduct is Meta’s censorship itself, rather than its policy of censoring. Based on the documents submitted and allegations made, that is a distinction without a difference.
RFK Jr. tried to argue that the ruling by Judge Doughty in Louisiana supports his position, but Judge Orrick wasn’t born yesterday and that he can actually read what the Supreme Court wrote in the Murthy decision rejecting these kinds of arguments.
The Murthy opinion makes my decision here straightforward. Murthy rejected Missouri’s factual findings and specifically explained that the Missouri evidence did not show that the federal government caused the content moderation decisions. Yet here, the plaintiffs rely on Missouri as their evidence that a state rule caused the defendants’ alleged censorship actions. Even if I accepted the vacated district court order as evidence here—which I do not—the Supreme Court has plainly explained why it does not support the plaintiffs’ argument.
Even though he notes that he doesn’t even need to go down this road, Judge Orrick also explains why the whole “state actor” argument is nonsense as well:
The plaintiffs’ theory is that Meta and the government colluded or acted jointly, or the government coerced Meta, to remove content related to Kennedy’s 2024 presidential campaign from Meta’s platforms. The problem with that theory is again the lack of evidence. The Missouri and Kennedy findings were rejected by the Supreme Court, as explained above. And they—and the interim report—suggest at most a relationship or communications between Meta and the government about removal of COVID-19 misinformation in 2020 and 2021. Even if the plaintiffs proved that Meta and the government acted jointly, or colluded, or that Meta was coerced by the government to remove and flag COVID-19 misinformation three years ago, that says nothing about Meta’s relationship and communications with the government in 2024. Nor does it suggest that Meta and the government worked together to remove pro-Kennedy content from Meta’s platforms.
Because of this, the plaintiffs fail to show likelihood of success on the merits—or serious questions going to the merits—for any of the three possible state action prongs. They do not provide evidence or allegations of a “specific[]” agreement between Meta and the government to specifically accomplish the goal of removing Kennedy content from Meta platforms. See Children’s Health, 2024 WL 3734422, at 5 (describing joint action test and collecting cases). Nor do they show that the government exercised coercive power or “significant encouragement” for Meta to remove Kennedy-related content in 2024. Id. at 9–10 (describing coercion test and finding that allegations about Congressmembers’ public criticism of COVID-19 misinformation on social media sites was insufficient to show government coerced platforms to remove it). And for similar reasons, the plaintiffs do not establish a “sufficiently close nexus” between the government and the removal of Kennedy-related content from Meta’s platforms. Id. at *5. Their First Amendment claim accordingly fails at step two of the state action inquiry. It is far from likely to succeed on the merits.
RFK Jr. also made a Voting Rights Act claim, that removing the documentary about him somehow interfered with people’s rights to vote for him. But the court notes that this argument is doomed by the fact that Meta noted that the blocking of links was an accident, which happens all the time:
The defendants point to compelling evidence that the video links were incorrectly automatically flagged as a phishing attack, a “not uncommon” response by its automated software to newly created links with high traffic flow. Oppo. 5–6 (citing Mehta Decl. Ex. A ¶ 7). The defendants’ evidence shows that once the defendants were alerted to the problem, through channels set up specifically for that purpose, the links were restored, and the video was made (and is currently still) available on its platform. Mehta Decl. Ex. A. ¶¶ 4–8, Exs. M–Q. Though the plaintiffs say the removal of the video was an effort to coerce them to not urge people to vote for Kennedy, the defendants’ competing evidence shows that it was a technological glitch and that the plaintiffs were aware of this glitch because they reported the problem in the first place. And if the plaintiffs were aware that a tech issue caused the removal of the videos, with that “context” it would probably not be reasonable for them to believe the video links were removed in an effort to coerce or intimidate them.
The court is also not impressed by the argument that other people (not parties to the case) had accounts removed or limited for sharing support for RFK Jr. As the judge makes clear, RFK Jr. doesn’t get to sue someone over a claim that they intimidated someone else (for which there isn’t any actual evidence anyway).
Third, the plaintiffs submit evidence that other peoples’ accounts were censored, removed, or threatened with removal when they posted any sort of support for Kennedy and his candidacy. See, e.g., Repl. 1:13–24; [Dkt No. 29-1] Exs. A, B. The defendants fail to respond to these allegations in their opposition, but the reason for this failure seems obvious. Section 11(b) provides a private right of action for Person A where Person B has intimidated, threatened, or coerced Person A “for urging or aiding any person to vote.” 52 U.S.C.A. § 10307(b). It does not on its face, or in any case law I found or the parties cite, provide a private right of action for Person C to sue Person B for intimidating, threatening, or coercing Person A “for urging or aiding any person to vote.” Id. Using that example, the three plaintiffs would be “Person C.” Their evidence very well might suggest that Meta is censoring other users’ pro-Kennedy content. But those users are not plaintiffs in this case and are not before me now.
Importantly, the plaintiffs had plenty of time and opportunity to add any of those affected users as new plaintiffs in this case, as they added Reed Kraus between filing the initial complaint and filing the AC and current motion. But they did not do so. Nor do they allege or argue that AV24 has some sort of organizational or third-party standing to assert the claims of those affected users. And while they seem to say that Kennedy himself is affected because that evidence shows Meta users are being coerced or threatened for urging people to vote for him, the effect on the candidate is not what § 11(b) protects. Accordingly, this evidence does not support the plaintiffs’ assertions. The plaintiffs, therefore, fail to counter the compelling evidence and reasons that the defendants identify in explanation for the alleged censorship.
More critically, the plaintiffs do not deny the defendants’ portrayal of and reasons for the defendants’ actions. The plaintiffs fail to incorporate those reasons into their assessment of how a “reasonable” recipient of Meta’s communications would interpret the communications in “context.” See Wohl III, 661 F. Supp. 3d at 113. Based on the evidence provided so far, a reasonable recipient of Meta’s communications would be unlikely to view them as even related to voting, let alone as coercing, threatening, or intimidating the recipient with respect to urging others to vote.
Towards the end of the ruling, the court finally gets to Section 230 and notes that the case is probably going nowhere even without everything earlier, because Section 230 makes Meta immune from liability for its moderation actions. However, the case didn’t hinge on that because neither side really went deep on the 230 arguments.
As for the other RFK Jr. case, I had forgotten that he had also sued Google/YouTube over its moderation efforts. At the end of last month, the Ninth Circuit also upheld a lower court ruling on that case in an unpublished four-page opinion where the three-judge panel made quick work of the nonsense lawsuit:
Google asserts that it is a private entity with its own First Amendment rights and that it removed Kennedy’s videos on its own volition pursuant to its own misinformation policy and not at the behest of the federal government. Kennedy has not rebutted Google’s claim that it exercised its independent editorial choice in removing his videos. Nor has Kennedy identified any specific communications from a federal official to Google concerning the removed Kennedy videos, or identified any threatening or coercive communication, veiled or otherwise, from a federal official to Google concerning Kennedy. As Kennedy has not shown that Google acted as a state actor in removing his videos, his invocation of First Amendment rights is misplaced. The district court’s denial of a preliminary injunction is AFFIRMED.
If RFK Jr. intends to appeal the latest Meta ruling (and given the history of his frivolous litigation, the chances seem quite high that he will), the Ninth Circuit might want to just repurpose this paragraph and swap out the “Google” for “Meta” each time.
Now, if only the Fifth Circuit would learn a lesson or two from the Ninth Circuit (or the Supreme Court), we could finally dispense with the one case that ridiculously went in RFK Jr.’s favor.
Filed Under: 1st amendment, 9th circuit, content moderation, free speech, jed rubenfeld, rfk jr., state actor, voting rights act, william orrick
Companies: google, meta, youtube
Police Union Says Fuck Officers, Law And Order; Tosses Endorsement To Donald Trump
from the we're-really-just-lobbyists dept
The Fraternal Order of Police isn’t here to protect cops from anything but accountability. It doesn’t actually care about the rank-and-file, not when it can leverage its power to secure even more power. It doesn’t care about law and order, despite being the definitive figurehead of that ideal.
No, the FOP is — and has always been — about securing its own power while providing a source for comments on police misconduct that journalists should never consider utilizing. It’s pure garbage. It’s the reason cops continue to dodge accountability. And it’s the reason law-and-order candidates keep getting elected, even when it’s clear literally anyone else would be a better option.
The FOP claims to be the voice of the rank-and-file. But the only officers it truly speaks for are the worst officers still capable of paying union dues. The presidential election is less than two months away. And, speaking only for itself, the Fraternal Order of Police has given its support to a candidate who cheered on supporters who openly assaulted police officers en route to invading a federal building.
Patrick Yoes, National President of the Fraternal Order of Police, announced today that the members of the FOP voted to endorse Donald J. Trump for President of the United States.
“Public safety and border security will be important issues in the last months of this campaign,” Yoes said. “Our members carefully considered the positions of the candidates on the issues and there was no doubt—zero doubt—as to who they want as our President for the next four years: Donald J. Trump.”
Gross. That’s a terrible decision by the FOP. Worse than that is Patrick Yoes’ claim this is representative of the view of a majority of officers who are union members. This is just the view of those leading FOP chapters. The rank-and-file aren’t exactly happy the FOP has endorsed a candidate who continues to unequivocally support the people who attacked their fellow officers during the raid of the Capitol building.
Durham County Sheriff Clarence Birkhead is joining over 100 law enforcement officials nationwide in condemning FOP’s endorsement. He said he is endorsing Democratic nominee Kamala Harris.
In its endorsement announcement, FOP called Trump a partner and a leader, saying the organization stands with him.
During a Harris-Walz press call, Birkhead called out the group for supporting a convicted felon over a former prosecutor. He said she “understands the needs of the communities like mine, and has always had our back.”
That’s not exactly welcome news for Kamala Harris, who has been trying to downplay her former career as a prosecutor. But, in all fairness, Harris never expressed support for criminals who attacked Capitol police officers. And, despite her past efforts to jail as many people as possible, she’s never suggested insurrectionists should be pardoned, much less hailed as “hostages” of an unfair government. That’s Trump’s territory. And despite his history of supporting lawlessness (and violating laws!), the top cop shop in the nation says Trump’s their man.
This endorsement shits on a ton of cops, especially those who had to deal with the aftermath of Trump’s refusal to engage in a peaceful transition of power in 2020. Here’s how the union, that claims to speak for hundreds of thousands of police officers, responded to the officers who were attacked during the January 6th Capitol raid:
As Officer Michael Fanone told Don Lemon, the FOP didn’t even reach out to him after the attack, despite his appearances on CNN. Fanone said that he was the one who had to reach out to the FOP about the incident. A whopping 140 members of the Capitol Police and D.C. Metro Police were injured that day and some are still out on medical leave. Others will never be able to work in their jobs again.
“I finally picked up the phone and called the president of the national FOP, Patrick Yoes and, and described to him the displeasure I felt that there was no outreach being done not only to myself but to other officers,” said Fanone. “And I asked him to do a few things to make up for that lack of support, and he was unwilling to do any of them. I asked him to publicly denounce the 21 house Republicans that voted against the gold medal bill.”
Democrats had a bill that would given the highest congressional honor to the officers who risked their lives that day. The FOP refused to speak out in support of that or against the Republicans who opposed it.
This alone should make it clear the FOP is not so much a police union as it is a political action committee. It doesn’t care about officers who are attacked by people who slap thin blue line bumper stickers on their trucks before mouthing off at city council meetings about their refusal to comply to with gun possession laws. The FOP is in business for itself, but it continues to make all officers pay for things they may not actually support. That sounds more like a political entity than a worker-focused entity that does everything it can to ensure all union members are treated fairly and are well-represented by their local union leaders.
This endorsement has only encouraged Donald Trump’s police state desires. Once again, Trump is promising to bring back things that have never worked because, if nothing else, they’ll result in more poor minorities being arrested and jailed.
Trump said if he wins in November, he would sign legislation to “strengthen protections for police officers” and would “crack down on Marxist prosecutors,” citing district attorneys in Philadelphia and Los Angeles specifically.
He also called for a return to “proven crime fighting methods, including stop and frisk and broken windows policing.” Trump has previously praised the use of stop and frisk policing and its use in New York City under former Mayor Rudy Giuliani (R).
These are the policies that were in place during the period where violent crime rates were at their highest. To say they didn’t work undersells the actual level of failure. We’re still in a period of historically low crime rates and yet we’re constantly being told by people like Trump and his supporters that America has reached record levels of lawlessness — something that can only be brought under control by expanding police powers and further limiting the rights of Americans. The support of the FOP only encourages this sort of thinking.
These two garbage entities were made for each other. They want more power and less accountability. And they claim they’re doing this for the little people — rank-and-file officers and/or millions of American voters. But they’re only doing it for themselves. And they’re only doing it because it will allow them to return to the open bigotry and casual rights violations that have long defined policing in America.
Filed Under: donald trump, fraternal order of police, january 6
Daily Deal: Babbel Language Learning (All Languages)
from the good-deals-on-cool-stuff dept
You probably already know the benefits of learning a language, so let’s focus on the app. Right off the bat, let’s be clear about one thing: When we say “app” we don’t mean that you’re limited to using Babbel on your phone. You can use Babbel on desktop, too, and your progress is synchronized across devices. Want to practice where you won’t have Wi-Fi? Download lessons before you head out, and you’ll be good to go. However you choose to access your 10K+ hours of online language education, you’ll be able to choose from 14 languages. Want to try your hand at all of them? Knock yourself out — you’ll have a lifetime to get it done. And you can tackle one or all in 10-to-15-minute bite-sized lessons, so there’s no need to clear hours of your weekend to gain real-life conversation skills. It’s on sale for $139.97 for the next 5 days only.
Note: The Techdirt Deals Store is powered and curated by StackCommerce. A portion of all sales from Techdirt Deals helps support Techdirt. The products featured do not reflect endorsements by our editorial team.
Filed Under: daily deal
Utah’s ‘Protect The Kids Online!’ Law Rejected By Court
from the utah-does-it-again dept
Over the last few years, politicians in Utah have been itching to pass terrible internet legislation. Some of you may forget that in the earlier part of the century, Utah became somewhat famous for passing absolutely terrible internet laws that the courts then had to clean up. In the last few years, it’s felt like other states have passed Utah, and maybe its lawmakers were getting a bit jealous in losing their “we pass batshit crazy unconstitutional internet laws” crown.
So, two years ago, they started pushing a new round of such laws. Even as they were slamming social media as dangerous and evil, Utah Governor Spencer Cox proudly signed the new law, streaming it on all the social media sites he insisted were dangerous. When Utah was sued by NetChoice, the state realized that the original law was going to get laughed out of court and they asked for a do-over, promising that they were going to repeal and replace the law with something better. The new law changed basically nothing, though, and an updated lawsuit (again by NetChoice) was filed.
The law required social media companies to engage in “age assurance” (which is just a friendlier name for age verification, but still a privacy nightmare) and then restrict access to certain types of content and features for “minor accounts.”
Cox also somewhat famously got into a fight on ExTwitter with First Amendment lawyer Ari Cohn. When Cohn pointed out that the law clearly violates the First Amendment, Cox insisted: “Can’t wait to fight this lawsuit. You are wrong and I’m excited to prove it.” When Cohn continued to point out the law’s flaws, Cox responded “See you in court.”
In case you’re wondering how the lawsuit is going, last night Ari got to post an update:
The law is enjoined. The court found it to likely be unconstitutional, just as Ari and plenty of other First Amendment experts expected. This case has been a bit of a roller coaster, though. A month and a half ago, the court said that Section 230 preemption did not apply to the case. The analysis on that made no sense. As we just saw, a court in Texas threw out a very similar law and said that since it tried to limit how sites could moderate content, it was preempted by Section 230. But, for a bunch of dumb reasons, the judge here, Robert Shelby, argued that the law wasn’t actually trying to impact content moderation (even though it clearly was).
But, that was only part of the case. The latest ruling found that the law almost certainly violates the First Amendment anyway:
NetChoice’s argument is persuasive. As a preliminary matter, there is no dispute the Act implicates social media companies’ First Amendment rights. The speech at issue in this case— the speech social media companies engage in when they make decisions about how to construct and operate their platforms—is protected speech. The Supreme Court has long held that “[a]n entity ‘exercis[ing] editorial discretion in the selection and presentation’ of content is ‘engage[d] in speech activity’” protected by the First Amendment. And this July, in Moody v. NetChoice, LLC, the Court affirmed these First Amendment principles “do not go on leave when social media are involved.” Indeed, the Court reasoned that in “making millions of . . . decisions each day” about “what third-party speech to display and how to display it,” social media companies “produce their own distinctive compilations of expression.”
Furthermore, following on the Supreme Court’s ruling earlier this year in Moody about whether or not the entire law can be struck down on a “facial” challenge, the court says “yes” (this issue has recently limited similar rulings in Texas and California):
NetChoice has shown it is substantially likely to succeed on its claim the Act has “no constitutionally permissible application” because it imposes content-based restrictions on social media companies’ speech, such restrictions require Defendants to show the Act satisfies strict scrutiny, and Defendants have failed to do so.
Utah tries to argue that this law is not about speech and content, but rather about conduct and “structure,” as California did in challenges to its “kids code” law. The court is not buying it:
Defendants respond that the Definition contemplates a social media service’s “structure, not subject matter.” However, Defendants’ argument emphasizes the elements of the Central Coverage Definition that relate to “registering accounts, connecting accounts, [and] displaying user-generated content” while ignoring the “interact socially” requirement. And unlike the premises-based distinction at issue in City of Austin, the social interaction-based distinction does not appear designed to inform the application of otherwise content-neutral restrictions. It is a distinction that singles out social media companies based on the “social” subject matter “of the material [they] disseminate[].” Or as Defendants put it, companies offering services “where interactive, immersive, social interaction is the whole point.”
The court notes that Utah seems to misunderstand the issue, and finds the idea that this law is content neutral to be laughable:
Defendants also respond that the Central Coverage Definition is content neutral because it does not prevent “minor account holders and other users they connect with [from] discuss[ing] any topic they wish.” But in this respect, Defendants appear to misunderstand the essential nature of NetChoice’s position. The foundation of NetChoice’s First Amendment challenge is not that the Central Coverage Definition restricts minor social media users’ ability to, for example, share political opinions. Rather, the focus of NetChoice’s challenge is that the Central Coverage Definition restricts social media companies’ abilities to collage user-generated speech into their “own distinctive compilation[s] of expression.”
Moreover, because NetChoice has shown the Central Coverage Definition facially distinguishes between “social” speech and other forms of speech, it is substantially likely the Definition is content based and the court need not consider whether NetChoice has “point[ed] to any message with which the State has expressed disagreement through enactment of the Act.”
Given all that, strict scrutiny applies, and there’s no way this law passes strict scrutiny. The first prong of the test is whether or not there’s a compelling state interest in passing such a law. And even though it’s about the moral panic of kids on the internet, the court says there’s a higher bar here. Because we’ve done this before, with California trying to regulate video games, which the Supreme Court struck down fourteen years ago:
To satisfy this exacting standard, Defendants must “specifically identify an ‘actual problem’ in need of solving.” In Brown v. Entertainment Merchants Association, for example, the Supreme Court held California failed to demonstrate a compelling government interest in protecting minors from violent video games because it lacked evidence showing a causal “connection between exposure to violent video games and harmful effects on children.” Reviewing psychological studies California cited in defense of its position, the Court reasoned research “show[ed] at best some correlation between exposure to violent entertainment” and “real-world effects.” This “ambiguous proof” did not establish violent videogames were such a problem that it was appropriate for California to infringe on its citizens’ First Amendment rights. Likewise, the Court rejected the notion that California had a compelling interest in “aiding parental authority.” The Court reasoned the state’s assertion ran contrary to the “rule that ‘only in relatively narrow and well-defined circumstances may government bar public dissemination of protected materials to [minors].’”
While there’s lots of screaming and yelling about how social media is bad for kids’ mental health, as we directly told Governor Cox, the evidence just doesn’t support the claim. The court seems to recognize that the claims are a lot of hot air as well. Indeed, Utah submitted the Surgeon General’s report as “proof,” which apparently they didn’t even read. As we noted, contrary to the media reporting on that report, it contained a very nuanced analysis that does not show any causal harms to kids from social media.
The judge absolutely noticed that.
First, though the court is sensitive to the mental health challenges many young people face, Defendants have not provided evidence establishing a clear, causal relationship between minors’ social media use and negative mental health impacts. It may very well be the case, as Defendants allege, that social media use is associated with serious mental health concerns including depression, anxiety, eating disorders, poor sleep, online harassment, low self-esteem, feelings of exclusion, and attention issues. But the record before the court contains only one report to that effect, and that report—a 2023 United States Surgeon General Advisory titled Social Media and Youth Mental Health—offers a much more nuanced view of the link between social media use and negative mental health impacts than that advanced by Defendants. For example, the Advisory affirms there are “ample indicators that social media can . . . have a profound risk of harm to the mental health and well-being of children and adolescents,” while emphasizing “robust independent safety analyses of the impact of social media on youth have not yet been conducted.” Likewise, the Advisory observes there is “broad agreement among the scientific community that social media has the potential to both benefit and harm children and adolescents,” depending on “their individual strengths and vulnerabilities, and . . . cultural, historical, and socio-economic factors.” The Advisory suggests social media can benefit minors by “providing positive community and connection with others who share identities, abilities, and interest,” “provid[ing] access to important information and creat[ing] a space for self-expression,” “promoting help-seeking behaviors[,] and serving as a gateway to initiating mental health care.”
The court is also not at all impressed by a declaration Utah provided by Jean Twenge, who is Jonathan Haidt’s partner-in-crime in pushing the baseless moral panic narrative about kids and social media.
Moreover, a review of Dr. Twenge’s Declaration suggests the majority of the reports she cites show only a correlative relationship between social media use and negative mental health impacts. Insofar as those reports support a causal relationship, Dr. Twenge’s Declaration suggests the nature of that relationship is limited to certain populations, such as teen girls, or certain mental health concerns, such as body image.
Then the court points out (thank you!) that kids have First Amendment rights too:
Second, Defendants’ position that the Act serves to protect uninformed minors from the “risks involved in providing personal information to social media companies and other users” ignores the basic First Amendment principle that “minors are entitled to a significant measure of First Amendment Protection.” The personal information a minor might choose to share on a social media service—the content they generate—is fundamentally their speech. And the Defendants may not justify an intrusion on the First Amendment rights of NetChoice’s members with, what amounts to, an intrusion on the constitutional rights of its members’ users…
Furthermore, Utah fails to meet the second prong of strict scrutiny, that the law be “narrowly tailored.” Because it’s not:
To begin, Defendants have not shown the Act is the least restrictive option for the State to accomplish its goals because they have not shown existing parental controls are an inadequate alternative to the Act. While Defendants present evidence suggesting parental controls are not in widespread use, their evidence does not establish parental tools are deficient. It only demonstrates parents are unaware of parental controls, do not know how to use parental controls, or simply do not care to use parental controls. Moreover, Defendants do not indicate the State has tried, or even considered, promoting “the diverse supervisory technologies that are widely available” as an alternative to the Act. The court is not unaware of young people’s technological prowess and potential to circumvent parental controls. But parents “control[] whether their minor children have access to Internet-connected devices in the first place,” and Defendants have not shown minors are so capable of evading parental controls that they are an insufficient alternative to the State infringing on protected speech.
Also, this:
Defendants do not offer any evidence that requiring social media companies to compel minors to push “play,” hit “next,” and log in for updates will meaningfully reduce the amount of time they spend on social media platforms. Nor do Defendants offer any evidence that these specific measures will alter the status quo to such an extent that mental health outcomes will improve and personal privacy risks will decrease
The court also points out that the law targets social media only, and not streaming or sports apps, but if it was truly harmful, then the law would have to target all of those other apps as well. Utah tried to claim that social media is somehow special and different than those other apps, but the judge notes that they provide no actual evidence in support of this claim.
But Defendants simply do not offer any evidence to support this distinction, and they only compare social media services to “entertainment services.” They do not account for the wider universe of platforms that utilize the features they take issue with, such as news sites and search engines. Accordingly, the Act’s regulatory scope “raises seriously doubts” about whether the Act actually advances the State’s purported interests.
The court also calls out that NetChoice member Dreamwidth, run by the trust & safety expert known best online as @rahaeli, proves how stupid and mistargeted this law is:
Finally, Defendants have not shown the Act is not seriously overinclusive, restricting more constitutionally protected speech than necessary to achieve the State’s goals. Specifically, Defendants have not identified why the Act’s scope is not constrained to social media platforms with significant populations of minor users, or social media platforms that use the addictive features fundamental to Defendants’ well-being and privacy concerns. NetChoice member Dreamwidth, “an open source social networking, content management, and personal publishing website,” provides a useful illustration of this disconnect. Although Dreamwidth fits the Central Coverage Definition’s concept of a “social media service,” Dreamwidth is distinguishable in form and purpose from the likes of traditional social media platforms—say, Facebook and X. Additionally, Dreamwidth does not actively promote its service to minors and does not use features such as seamless pagination and push notification.
The court then also notes that if the law went into effect, companies would face irreparable injury, given the potential fines in the law.
This harm is particularly concerning given the high cost of violating the Act—$2,500 per offense—and the State’s failure to promulgate administrative rules enabling social media companies to avail themselves of the Act’s safe harbor provision before it takes effect on October 1, 2024.
Some users also sued to block the law, and the court rejected that request as there is no clear redressable injury for those plaintiffs yet, and thus they have no standing to sue at this point. That could have changed after the law started to be enforced, but thanks to the injunction from the NetChoice part, the law is not going into effect.
Utah will undoubtedly waste more taxpayer money and appeal the case. But, so far, these laws keep failing in court across the country. And that’s great to see. Kids have First Amendment rights too, and one day, our lawmakers should start to recognize that fact.
Filed Under: 1st amendment, age assurance, age verification, content moderation, kids, protect the children, robert shelby, social media, utah
Companies: netchoice
GOP Holds Show Hearing To Complain About Providing Affordable Broadband To Poor People
from the poor-people-have-it-too-easy dept
Over the next six months, states are poised to receive more than $42.5 billion in taxpayer subsidies to help fund broadband rollouts around the country. A lot of this money is getting dumped into the laps of big telecom monopolies with a lousy track record of follow through. But a lot of it is also going to transformative efforts by cooperatives, municipalities, and community-owned networks.
Republicans voted against COVID relief legislation (ARPA) and the infrastructure bill (IIJA), but they still like taking credit for both bills’ broadband funding efforts with local constituents (who’s going to correct locals’ perceptions, Sinclair broadcasting?).
At the same time, Republicans are holding a series of hearings because they’re very upset about a small component of the infrastructure bill’s BEAD (Broadband Equity And Deployment) program: namely that it (gasp!) requires ISPs to try and provide a basic, affordable broadband tier for low-income Americans.
Regional telecom monopolies like AT&T and Comcast are always worried that any government effort to force them to actually try or compete on price could snowball into more such efforts to fix our broken telecom markets (it never does due to widespread corruption, but this is their persistent worry).
So they’ve convinced Republican House Energy and Commerce Committee Chair Cathy McMorris Rodgers and Communications and Technology Subcommittee Chair Bob Latta to hold a series of silly hearings this week that will focus on attacking the low-income provision of the infrastructure bill.
The first such hearing is slated for this week, and the witness list doesn’t include anybody from the Biden Administration. Its underlying claim will be that requiring that states ensure poor people can afford the broadband access (they’re funding with their own taxpayer dollars) is akin to “rate regulation”:
“The [Infrastructure Investment and Jobs Act of 2021] prohibits the NTIA from regulating the rates charged for broadband service. At the same time, NTIA requires states receiving BEAD funds to define a low-cost broadband service option. NTIA is responsible for reviewing and approving these low-cost options and has used this requirement as a way to regulate rates,” the staff memo said. “The Committee considers these rate regulated approvals to be a violation of the IIJA’s rate regulation prohibition.”
Telecoms have always vehemently opposed any effort to address the high prices and spotty access caused by two major things: unchecked telecom consolidation and monopolization, and the corruption that protects it. It doesn’t matter that our regulators are generally too cowed to actually do that; any effort to address runaway costs is portrayed as radical extremism, and “rate regulation” is always thrown around as this menacing bogeyman that never actually comes to town in any real way.
The law in question, IIJA, delegates fund management to the states, which should start receiving money this fall. It also requires that providers that take taxpayer money provide at least one “low-cost broadband service option for eligible subscribers.” But the law also says the NTIA may not “regulate the rates charged for broadband service.” At a hearing last May, NTIA Administrator Alan Davidson put it this way:
“The statute requires that there be a low-cost service option. We do not believe the states are regulating rates here. We believe that this is a condition to get a federal grant. Nobody’s requiring a service provider to follow these rates, people do not have to participate in the program.”
AT&T and friends want to hoover up billions in taxpayer dollars with no meaningful conditions on how those funds are utilized. So Republicans are poised to put on a show this week (with the press’ help, I’m sure) about how unfair it is that AT&T and Comcast have to offer a $20, 25 Mbps service tier (which costs them virtually nothing to provide) to families that qualify for low-income lunch programs.
I suspect that whether Republicans succeed or not at stripping away the low-cost provisions (which were never going to be enforced with any zeal anyway), you can absolutely be certain they’ll take singular credit for the bill’s broadband deployments, despite fighting the program every step of the way.
Filed Under: BEAD, bob latta, broadband, cathy mcmorris rodgers, deployment, high speed internet, infrastructure bill, telecom
NO FAKES – A Dream For Lawyers, A Nightmare For Everyone Else
from the no-fakes,-no-good dept
Performers and ordinary humans are increasingly concerned that they may be replaced or defamed by AI-generated imitations. We’re seeing a host of bills designed to address that concern – but every one just generates new problems. Case in point: the NO FAKES Act. We flagged numerous flaws in a “discussion draft” back in April, to no avail: the final text has been released, and it’s even worse.
Under NO FAKES, any human person has the right to sue anyone who has either made, or made available, their “digital replica.” A replica is broadly defined as “a newly-created, computer generated, electronic representation of the image, voice or visual likeness” of a person. The right applies to the person themselves; anyone who has a license to use their image, voice, or likeness; and their heirs for up to 70 years after the person dies. Because it is a federal intellectual property right, Section 230 protections – a crucial liability shield for platforms and anyone else that hosts or shares user-generated content—will not apply. And that legal risk begins the moment a person gets a notice that the content is unlawful, even if they didn’t create the replica and have no way to confirm whether or not it was authorized, or have any way to verify the claim. NO FAKES thereby creates a classic “hecklers’ veto”: anyone can use a specious accusation to get speech they don’t like taken down.
The bill proposes a variety of exclusions for news, satire, biopics, criticism, etc. to limit the impact on free expression, but their application is uncertain at best. For example, there’s an exemption for use of a replica for a “bona fide” news broadcast, provided that the replica is “materially relevant” to the subject of the broadcast. Will citizen journalism qualify as “bona fide”? And who decides whether the replica is “materially relevant”?
These are just some of the many open questions, all of which will lead to full employment for lawyers, but likely no one else, particularly not those whose livelihood depends on the freedom to create journalism or art about famous people.
The bill also includes a safe harbor scheme modelled on the DMCA notice and takedown process. To stay within the NO FAKES safe harbors, a platform that receives a notice of illegality must remove “all instances” of the allegedly unlawful content—a broad requirement that will encourage platforms to adopt “replica filters” similar to the deeply flawed copyright filters like YouTube’s Content I.D. Platforms that ignore such a notice can be on the hook just for linking to unauthorized replicas. And every single copy made, transmitted, or displayed is a separate violation incurring a $5000 penalty – which will add up fast. The bill does throw platforms a not-very-helpful-bone: if they can show they had an objectively reasonable belief that the content was lawful, they only have to cough up $1 million if they guess wrong.
All of this is a recipe for private censorship. For decades, the DMCA process has been regularly abused to target lawful speech, and there’s every reason to suppose NO FAKES will lead to the same result.
What is worse, NO FAKES offers even fewer safeguards for lawful speech than the DMCA. For example, the DMCA includes a relatively simple counter-notice process that a speaker can use to get their work restored. NO FAKES does not. Instead, NO FAKES puts the burden on the speaker to run to court within 14 days to defend their rights. The powerful have lawyers on retainer who can do that, but most creators, activists, and citizen journalists do not.
NO FAKES does include a provision that, in theory, would allow improperly targeted speakers to hold notice senders accountable. But they must prove that the lie was “knowing,” which can be interpreted to mean that the sender gets off scot-free as long as they subjectively believes the lie to be true, no matter how unreasonable that belief. Given the multiple open questions about how to interpret the various exemptions (not to mention the common confusions about the limits of IP protection that we’ve already seen), that’s pretty cold comfort.
These significant flaws should doom the bill, and that’s a shame. Deceptive AI-generated replicas can cause real harms, and performers have a right to fair compensation for the use of their likenesses, should they choose to allow that use. Existing laws can address most of this, but Congress should be considering narrowly-targeted and proportionate proposals to fill in the gaps.
The NO FAKES Act is neither targeted nor proportionate. It’s also a significant Congressional overreach—the Constitution forbids granting a property right in (and therefore a monopoly over) facts, including a person’s name or likeness.
The best we can say about NO FAKES is that it has provisions protecting individuals with unequal bargaining power in negotiations around use of their likeness. For example, the new right can’t be completely transferred to someone else (like a film studio or advertising agency) while the person is alive, so a person can’t be pressured or tricked into handing over total control of their public identity (their heirs still can, but the dead celebrity presumably won’t care). And minors have some additional protections, such as a limit on how long their rights can be licensed before they are adults.
Republished from the EFF’s Deeplinks blog.
Filed Under: ai, congress, digital replicas, generative ai, likeness rights, no fakes
Ordinary Meaning, Extraordinary Methods: Judge Newsom Digs Deeper Into AI In The Courtroom
from the what-is-the-ordinary-meaning-of-ordinary-meaning dept
Earlier this year, we wrote about how Judge Kevin Newsom, on the Eleventh Circuit Court of Appeals, had explored how ChatGPT might actually be useful for a particularly narrow use in a court. Specifically, in judging whether or not the “ordinary meaning” of a phrase matched with what a party in the court argued was the “ordinary meaning” of a phrase.
Newsom was quite thoughtful and careful in his analysis, highlighting the potential risks and limitations. However, he noted that given that ChatGPT is trained across a broad selection of content, it might actually be useful to see whether or not it agreed that the “ordinary meaning” of a particular term (in this case “landscaping”) covered how it was used in the case at hand.
Judge Newsom has continued to think about this and has now released another concurring opinion, which he notes is a sort of “sequel” to the original. Again, it explores an area where an LLM might actually help a judge puzzle through something. This case also involves an issue of “ordinary meaning” — this time about what “physical restraint” means.
Judge Newsom notes that having yet another case regarding an issue of ordinary meaning allowed him to tackle an issue that many people had raised regarding his original: which is that LLMs may give different results each time you ask.
The LLM research that I conducted in connection with this case, though, raised a question that I hadn’t squarely confronted in Snell, one that seemed worth flagging and exploring: What should we make of the fact that the models sometimes provide subtly different answers to the exact same question? I’ll confess that the variation initially spooked me, but it now (for reasons I’ll get into) seems not only unremarkable but perhaps even expected. And significantly from an ordinary-meaning perspective, it accurately reflects real people’s everyday speech patterns.
First, he notes that “physically restrained” may be even tougher to define than “landscaping” because it’s two words, and thus not something that is directly defined in a dictionary. The response to that (which is how the court actually ruled in this case) was to break up the term into its component parts, sort through the meaning of each word, and put it all back together.
But perhaps ChatGPT could provide a different approach, given that it is designed to “capture and reflect how real people ordinarily use both words and multi-word phrases.” Judge Newsom notes that he couldn’t help but experiment, and shares his query and response.
Me: “What is the ordinary meaning of ‘physically restrained’?”
ChatGPT: The ordinary meaning of “physically restrained” refers to the act of limiting or preventing someone’s movement by using physical force or some kind of device. This can involve holding someone down, using handcuffs, straps, or other mechanisms to restrict their ability to move freely. The term implies an active and physical intervention to control or confine a person’s movements.
Not wanting to trust it just because it agreed with his priors, he decided to also ask Anthropic’s Claude (which remains my favorite LLM):
For reasons I’ve explained before, though, I didn’t want to seize on ChatGPT’s explanation just because it matched my priors. See Snell, 102 F.4th at 1225 (Newsom, J., concurring). So, I asked Claude3 the same question: “What is the ordinary meaning of ‘physically restrained’?” Reassuringly, Claude’s response largely mirrored ChatGPT’s:
The ordinary meaning of “physically restrained” generally refers to:
1. The act of using physical force or methods to limit or control someone’s movements or actions.
2. A state in which a person is prevented from moving freely or acting as they wish due to physical barriers or interventions. This can involve various forms of restraint, such as:
- Holding someone back with one’s hands or body
- Using devices like handcuffs, straps, or restraining jackets
- Confining someone to a specific space, like a locked room
The term is often used in contexts related to law enforcement, medical settings, or situations where someone’s freedom of movement needs to be restricted for safety or security reasons. However, it can also be used in more general situations where physical limitation is applied.
But then he tried again and noticed it gave a different response:
Then, for reasons I don’t specifically recall—but that can presumably be chalked up to a “better safe than sorry” instinct—I asked Claude the exact same question again. To my surprise, its answer the second time around was ever-so-slightly different:
The ordinary meaning of “physically restrained” typically refers to:
Using physical means to limit or prevent someone’s movement or actions.
This usually involves:
- Holding someone back
- Using restraining devices (e.g. handcuffs)
- Confining someone to a limited space
It implies direct physical intervention to restrict a person’s freedom of movement.
As he points out, while the underlying point is the same, the actual answers are slightly different:
Hmm. Claude’s responses were basically the same—but not exactly, not verbatim. The substance seemed pretty much identical, but the style and structure were just a smidgen off. Most obviously, the first was longer and more detailed than the second. For instance, unlike the first response, the second didn’t include a “state in which” explanation. In the “holding someone back” bullet, the first specified a means—“with one’s hands or body”—while the second didn’t. Similarly, in the “devices” bullet, the first referred to handcuffs, straps, and restraining jackets as examples, while the second referred only to handcuffs. Finally, the first response concluded with a short paragraph about the various “contexts” in which the phrase “physically restrained” is often used—law enforcement, medicine, etc.—while the second closed with a more generic statement that the phrase “implies direct physical intervention to restrict a person’s freedom of movement.”
Judge Newsom wondered if this created problems for his idea of using LLMs in this manner. Specifically, he worried if this stochastic output meant that the LLM wasn’t “accurately communicating” what its corpus of knowledge suggested was an “ordinary meaning.”
So he did more experimentation. He ran the same queries ten times each on ChatGPT, Claude, and Gemini (using the freely available models of each). With 30 results across three different engines, he wondered if he might be able to learn something, including whether or not he could trust these kinds of answers if they all seemed to resolve to a similar underlying meaning.
Again reassuringly, the 30 results I received—10 apiece from each of the three leading LLMs—largely echoed the initial response that I got from ChatGPT. If you’re interested in the nitty gritty, all the responses are available in the Appendix. But here’s the gist: When defining “physically restrained,” the models all tended to emphasize “physical force,” “physical means,” or “physical barriers.” ChatGPT and Claude specifically used one (or more) of those phrases in every one of their responses. For whatever reason, Gemini was a little different. It didn’t invariably employ one of those terms explicitly, but even when it didn’t, the concept of what I’ll call corporeality (via either human touch or a tangible object) pervaded and tied together its example-laden answers.
To be sure, the models’ responses exhibited some minor variations in structure and phrasing. ChatGPT’s answers, for example, tended to fluctuate in length by a sentence or two. For its part, Claude altered the number of examples it provided from one response to the next. But for reasons I’ll explain in the next part, these subtle, marginal divergences were probably (and should have been) expected. Far more importantly, I think, the responses did coalesce, substantively, around a common core—there was an objectively verifiable throughline. For our purposes, what matters is that the LLMs consistently defined the phrase “physically restrained” to require the application of tangible force, either through direct bodily contact or some other device or instrument. And that, again, squares comfortably with the results obtained through the traditional, dictionary-driven breaking-and-repiecing method.
Newsom concludes that all of this makes him less worried about the lack of direct repeatability among engines, because, if anything, it makes it seem almost more human.
So, what to make of the slight variations among the answers that the models returned in response to my query? For present purposes, I think there are two important points. First, there’s a technical explanation for the variation, which, upon reflection, doesn’t much concern me—or, upon further reflection, even much surprise me. Second, there is, upon even further reflection, a sense in which the substantively-identical-and-yet-marginally-different answers (perhaps ironically) underscore the models’ utility in the ordinary-meaning analysis—namely, in that they pretty closely mimic what we would expect to see, and in fact do see, in everyday speech patterns.
As he explains later in the concurrence, you would expect the same variations if you just asked a bunch of people:
Remember, our aim is to discern “ordinary meaning.” Presumably, the ideal gauge of a word’s or phrase’s ordinary meaning would be a broad-based survey of every living speaker of American English—totally unrealistic, but great if you could pull it off. Imagine how that experiment would go: If you walked out onto the street and asked all umpteen million subjects, “What is the ordinary meaning of ‘physically restrained’?”, I think I can confidently guarantee that you would not get the exact same answer spit back at you verbatim over and over and over. Instead, you’d likely get a variety of responses that differed around the margins but that, when considered en masse, revealed a common core. And that common core, to my way of thinking, is the ordinary meaning.
Thus, the “problem” of variability in answers might not even be really a problem at all.
So, as it turns out, the very thing that had initially given me pause—namely, that the LLMs were returning subtly different responses to the same question—has instead given me (more) hope that the models have something significant to offer the interpretive enterprise. The fact is, language is an organic thing, and like most organic things, it can be a little messy. So too, unsurprisingly, are our efforts to capture its ordinary meaning. Because LLMs are trained on actual individuals’ uses of language in the real world, it makes sense that their outputs would likewise be less than perfectly determinate—in my experience, a little (but just a little) fuzzy around the edges. What’s important, though—and I think encouraging—is that amidst the peripheral uncertainty, the LLMs’ responses to my repeated queries reliably revealed what I’ve called a common core.
Before people freak out, he’s quite clear that he’s not suggesting this replace human judgment or that this is the be-all end-all of any such “ordinary meaning” determination:
A final coda: No one should mistake my missives for a suggestion that AI can bring scientific certainty to the interpretive enterprise. As I’ve been at pains to emphasize, I’m not advocating that we give up on traditional interpretive tools—dictionaries, semantic canons, etc. But I do think—and increasingly so—that LLMs may well serve a valuable auxiliary role as we aim to triangulate ordinary meaning.
And he leaves himself open to the most human of responses:
Again, just my two cents. I remain happy to be shouted down.
Filed Under: 11th circuit, ai, chatgpt, claude, gemini, generative ai, kevin newsom, llms, ordinary meaning
Florida Gov’t Sending Cops To Voter’s Houses To Intimidate Pro-Choice Ballot Measure Petition Supporters
from the fiefdom-intensifies dept
Well, here’s even more ugliness from America’s penis. The state of Florida — currently presided over by a bootlicker wearing boot lifts — has decided the best way to protect its war on women is to harass and intimidate residents who support reproductive rights.
Florida is reportedly sending police officers to the homes of people who signed a petition supporting an abortion rights ballot initiative.
Florida Governor Ron DeSantis’s administration has begun investigating thousands of verified signatures that helped to put a state constitutional amendment protecting the right to abortion on the ballot in November. The amendment would overturn Florida’s current six-week abortion ban.
[…]
Multiple residents in Lee County have reported being visited by law enforcement following up on their petition signatures.
This is a bold [read: fascist] response to suspected petition fraud. Suspected fraud like this has almost always been handled in far less invasive fashion. At worst, it’s usually just a handful of subpoenas. In most cases, the state approaches the court to ask it to invalidate the ballot measure.
But in Florida, the new thing is sending investigators to people’s houses with a folder full of ballot supporters’ personal information. Here’s what Florida resident Isaac Menasche observed when a law enforcement officer showed up at his house to question him about his ballot measure signature.
“The experience left me shaken. What troubled me was he had a folder on me containing my personal information—about 10 pages. I saw a copy of my drivers license and copy of the petition I signed. It was obvious to me that a significant effort was exerted to determine if indeed I had signed the petition.”
If this seems melodramatic, ask yourself how you would respond if a law enforcement officer showed up at your house and started asking questions about your support for political issues and candidates — especially while flipping through some sort of dossier you’d rightfully be surprised to find the state had compiled on you.
This is definitely not normal behavior. State governments generally don’t handle questionable ballot initiatives this way. But Ron DeSantis doesn’t want his anti-abortion law undone by the will of the people. And that means the people supporting this measure must be oppressed, if not actually suppressed.
The petition for the ballot measure supporting abortion until viability cleared the 900,000 signature mark by over 100,000 signatures. With his petty fiefdom threatened, Governor DeSantis has weaponized the state’s law enforcement agencies to quell this particular dissent, as the Tampa Bay Times reports.
The officer’s visit appears to be part of a broad — and unusual — effort by Gov. Ron DeSantis’ administration to inspect thousands of already verified and validated petitions for Amendment 4 in the final two months before Election Day. The amendment would overturn Florida’s six-week abortion ban by proposing to protect abortion access in Florida until viability.
Since last week, DeSantis’ secretary of state has ordered elections supervisors in at least four counties to send to Tallahassee at least 36,000 petition forms already deemed to have been signed by real people.
Additionally, a spokesperson for the governor’s office said it would be referring this to the state’s Department of Law Enforcement. Whether or not that has happened yet is unclear. The DLE has so far refused to comment on this highly unusual move by the state government.
DeSantis and his supporters in the legislature aren’t willing to engage in fair fights. When opposition arises, DeSantis and his enablers are quick to suppress it. Case in point: the governor’s firing of district attorneys who opposed his unconstitutional mandates and tough-on-crime antics. Now, there’s this: blatant voter intimidation to salvage an abortion ban that a considerable number of state residents don’t support.
This isn’t outlier behavior. This is the Republican party in its current form. Donald Trump spent the weekend inciting violence if the election didn’t go his way and promising bloodshed to enforce his mass deportation plans. And we all know DeSantis and his office would never engage in tactics like these if the alleged fraudulent activity involved ballot measures or candidates DeSantis supports. This is just bog-standard evil. Unfortunately, it’s the banal evils that open the door for the more horrific evils — you know, things like genocide and mass imprisonment of the government’s enemies. This is just another Overton Window being opened to see what the masses are willing to tolerate from their government.
Filed Under: 1st amendment, democracy, florida, intimidation, police state, ron desantis
