Thread by @corbinkbarthold

1/ Have I gotten enough sleep to coherently live-tweet the Gonzalez v. Google #Section230 argument at SCOTUS?

Let's find out!

supremecourt.gov/oral_arguments…

2/ If you need a refresher on, or intro to, the case -- what Sec. 230 is, how it came to be, what's at stake in this case -- there are a thousand pieces floating around.

But this here's my nifty thread (😏) so here's my essay for Reason doing all that.

reason.com/2022/11/04/sec…

3/ And, FYI, after the argument I'll be doing a breakdown of the argument for @ProgressChamber, alongside @jess_miers, @CathyGellis, and @ericgoldman.

us06web.zoom.us/meeting/regist…

4/ Massive shoutout to @nsaadbembridge, @jess_miers, @Will_Duffield, and @ChrisMarchese9 -- who are the VERY FIRST PEOPLE IN LINE for the argument today. 😂

https://twitter.com/nsaadbembridge/status/1628010989785341952?s=20

5/ The great @AriCohn is taking ... a different approach.

https://twitter.com/AriCohn/status/1627889053126254594?s=20

6/ What will I be watching for during this argument?

#1 Above all: How much do the justices dig into stuff that's not at issue in this case?

Do they take J. Thomas, Sen. Hawley, Sen. Cruz, etc., up in exploring whether there's a "publisher" v. "distributer" distinction?

7/ #1 cont. Do they spend time talking about how 230 governs *taking down* content (conservatives' big bugbear), rather than *leaving it up* (the only issue here)?

Do they ask whether 230 is speaker or content discriminatory, or some kind of unconstitutional condition (😬)?

8/ #2 Do the justices think a line can be drawn between algorithmic recommendations & all the other organizing platforms do as publishers of their sites?

The US wants the Court to draw such a line -- but I expect the justices to give it a lot of trouble about where that line is.

/9 #3 Just how confused will the justices be about the petitioners' position?

I fully expect they'll *be* confused -- the petitioners have shifted around! Question is, will the Court ~help the pets get themselves straightened out? Or will the justices ~throw up their hands?

10/ #4 is related to #3: Will the justices be so annoyed with the petitioners' tactics -- pets even changed the question presented! -- that they float the idea of dismissing the case as improvidently granted?

https://twitter.com/corbinkbarthold/status/1626243487585955840?s=20

11/ @ChrisMarchese9 with the proof!

Been there, done that. So glad I didn't wake up before dawn to head to SCOTUS (but instead woke up ... before dawn ... bc of screaming kids ... and headed to my living room🤔).

https://twitter.com/ChrisMarchese9/status/1628028127862525952?s=20

12/ Anyway, #5, and finally: To what degree are the justices trying to square this case with Twitter v Taamnah and -- even more so -- @NetChoice v Paxton/Moody and 303 Creative v Elenis?

A lot of moving pieces right now for free speech and the internet!

13/ #5 cont. So, for example, some people think you could narrow 230 by drawing a distinction between "conduct" and "speech." You're protected *literally* for hosting speech -- but not for the ills the speech might cause in the world ...

https://twitter.com/JuliaAngwin/status/1627643783776030720?s=20

14/ #5 cont. But hey! Why stop there? If we're gonna play that game, let's just say that *all* of what platforms do is conduct, rather than speech!

That's what the Fifth Circuit did in upholding HB20, Texas's law that requires social media services to carry hate speech, etc.

15/ #5 cont. There are many crisscrossing issues like that in these cases. Press a line of logic here, & it might blow up in your face there.

For the US, Rumsfeld v FAIR might look like a great precedent in 303 Creative, but an awful one in Paxton. Well, wield it at your peril!

16/ Alright -- that's it by way of preview. On to the main event! Starts in about 10 minutes. Just time to get coffee and see what my toddlers are fighting about in the hallway ...

17/ OK! Hyper-caffeinated. Kids cruelly abandoned to their own little Lord of the Flies situation. Let's goooooo!

The Chief Justice says we'll now hear argument in case number 21-1333, Gonzalez v. Google ... Eric Schnapper to start us off for petitioners ...

18/ Schnapper goes straight to arguing that "publisher," in (c)(1), is used in a technical, common-law sense derived from defamation law --> argument that has the potential to wipe *all* recommendations from 230.

19/ He doubles down on his abandonment of treating this case about "recommendations" as such -- what he initially wanted, in his cert. petition.

20/ Thomas opens the questioning by trying to differentiate different sorts of recommendation algorithms. Thomas to petitioners: GIVE ME A LINE TO DRAW.

21/ Schnapper: Did you know that there are these things called thumbnails?

22/ Schnapper: No Section 230 for thumbnails! 😬

23/ Thomas: let's talk about rice pilaf from Louisiana ... is that like an ISIS video? (??)

Basically, how do we separate these good algorithms we like from these bad algorithms we don't? (Are you a good witch, or a bad witch?)

24/ Roberts: If there's no focused algorithm -- no, uh, "terrorist" algorithm -- then YouTube's not responsible for it, right?

(Not bad for YouTube -- still not the greatest line of reasoning for, say, Yelp or Glassdoor.)

25/ Kagan: "This was a pre-algorithm statute." (Um ... sort of?) At any rate, algorithms are now "endemic to the internet." Does pets position send us down the road of having Section 230 not mean anything at all?

26/ Schnapper: Depends on what the algorithm is used for.

Well, yeah! If YouTube must know *everything* it recommends, and sort the good from the bad, that's basically eliminating 230 protection.

27/ Similar to what the EARN IT Act people do, in saying they just want to ditch "misuse" of encryption.

'We're not against [x]; we're just against the fundamental way [x] works!'

28/ Schnapper can't seem to decide whether he's arguing "not treated as a publisher," or "not third-party content." Sort of toggling between the two ...

29/ Alito: "I don't know where you're drawing the line. That's the problem." 😬

30/ Schnapper: "There'd have to be a cause of action for causing people to go look at the video."

???

So now we're limiting this 230 limitation to, basically, just Anti-Terrorism Act claims? That's new ...

31/ Jackson: I thought you were saying there should be some kind of liability for defective screening of content?

Schnapper: No, we're saying if you encourage people to go look at the content.

Um, does a newspaper "encourage" you to look at what it puts above the fold on A1?

32/ Jackson: Google's going to say "It's the same thing," recommendation versus presentation. (Yup.)

Schnapper: Guess we need to clarify the distinction we're drawing. (Yuuuuup.)

Doubles down again on the word "encouraging," like that solves something ...

33/ Sotomayor: you offer "paragraphs after paragraphs after paragraphs" saying they're liable for failing to take ISIS off their website. But now you seem to have abandoned that?

Schnapper: Yes.

34/ Soto + Schnap exchange: Nor is this case about networking. It's all about content.

Soto: So what part of what YouTube does counts as the aiding and abetting under the ATA? "The providers can be liable for what?"

Soto: it needs to be in your complaint!

Schnapper: $7j#4$p*@

35/ Schnapper: Falls back to the (absolutely arbitrary and illusory) distinction between content you ask for and content that's served up to you.

Soto: Not buying it. How do you get yourself from a neutral algorithm to intent? "How do you get there"?

36/ Schnapp: Uh, we'll talk about that tomorrow in Taamneh.

Roberts cuts in -- is YouTube akin to a bookstore recommending books?

Schnapp: flailing -- literally says he's worried where the question is heading ...

Roberts (roughly): Well, yeah, you should be!

37/ Schnapper: Tries to shift bookstore analogy to bookstore saying, in response to request, "Here's a book *we wrote*."

Yeah ... that's a totally different case.

38/ Thomas: trying to parse a distinction between "suggestions" and "recommendations" (?).

Schnapper: Yeah, we're not arguing that either. "Take the standards and apply them to what's going on." (Gee, *that's* helpful ...)

39/ Schnapper: "Is that responsive?"

Thomas: "It's responsive, but I don't understand."

Thomas: Where's the aiding and abetting?

Schnap: I'll answer that tomorrow.

Thomas: You're alleging mere inaction. That's not going to cut it.

Schnap: We'll get there tomorrow.

40/ Alito: I'm "completely confused" about "the argument you're making at the present time."

Ouch.

41/ Schnapper: Thumbnail is a joint creation.

Alito: So then YouTube is responsible for everything thumbnail it displays?

Schnap: Yup.

Alito: OK, so you're arguing that 230 protects nothing that's based on third-party content!

42/ Schnap: It's their business model.

Alito: So you're saying they shouldn't use thumbnails at all.

Schnap: gurgling noises.

43/ Soto: "Your position has gone further than I thought."

Question that begins "Assume I think you're wrong ..." (Never good!)

Schnap: "It all depends." Falling back on expectation that platforms magically separate all "good" from all "bad" material.

44/ Soto: "Assume you've lost" on your primary argument. Frames hypothetical in which platform and ISIS ~actively collude.

[some back and forth]

Soto: testing line where platform writes "discriminatory" algorithm ... (uh oh?)

Schnap, served softball: swings and misses.

45/ Kagan: serves up some hypos, with different degrees of "recommendation" from the platform. Ends with search engines.

Schnap: search engines protected.

Kagan: What!? The search engine served you up a recommendation!

Schnap: long silence ...

46/ Kagan: So even a search engine could be liable for its prioritization, right?

Schnap: incoherent response re: the EU fining a search engine for an antitrust violation (???)

47/ Kagan: We're a court, we shouldn't be figuring this stuff out. We're not internet experts. Lines are difficult to draw in this area. "Isn't this something for Congress to do?"

Schnap: Begs the question. "Apply the statute the way it was written."

48/ Schnap: A lot of stuff on the internet is highly desirable but doesn't fit within 230. Tough luck! No 230 protection for thumbnails.

(Well, for one thing, pretty sure we had thumbnails in 1996??)

49/ Gorsuch: I think what Google does in picking, choosing, analyzing, just makes it an ICS -- protected activity. Gotta do something more. Maybe we just kick it back to 9th Cir. to figure out what that "more" is?

Schnap: We'll, that's not our arg. (Dude, take a W!)

50/ Kavanaugh: The very thing that makes YouTube an interactive computer services is what, in your view, takes it outside 230. That makes no sense. Cites 230(f)(4).

Schnap: *butchers statute*

51/ Kav: Congress drafted a broad text. All the lower courts think it protects in this sort of situation. Shouldn't we stick with that reliance interest, and let Congress address any problem?

Schnap: Eh, narrowing 230'll be fine.

52/ Schnap: Most recommendations aren't actionable.

Note that he's basically saying: eviscerate 230, and it'll be fine because most causes of action -- aside from the ATA -- will lose.

That's the arg that like every losing plaintiff has ever made against 230.

53/ Barrett: 230 protects not only providers but users ...

YES! YES! YES!

Asks whether you can violate ATA with your Twitter account.

Well done.

54/ Barrett: On your theory, could I be liable for aiding and abetting when I retweet?

Schnap: LOUD NOISES.

Barrett: I don't understand, logically, why I'm not creating "new" content when I retweet. (🔥)

55/ Jackson: "I guess I'm thoroughly confused."

She thinks maybe Schnap is conflating 230 protection and ATA liability.

Oof ... but she then shifts into saying 230 is limited to overturning Stratton Oakmont?

56/ Schnapper then wastes time confirming that 230 has three elements, instead of running with the ball Jackson tried to hand him ...

... and that's it for Schnapper. Closes with a whimper.

57/ Next up: Malcom Stewart for the United States.

Goes straight to the bookstore example.

Bottom line: thinks there's no protection the moment there's a statement to the effect of "Check this content out."

58/ Thomas: I'm still confused.

Back and forth trying to distinguish recommendation versus presentation -- exactly the line the US never really drew in its brief.

59/ Sounds like Stewart would take all recs out of 230.

If you do a search, and there's a typo, and the search engine still gives you a solid result, then ... there's no 230 protection!?

60/ Kagan: You can't present content without making choices. In every case with content, there's a choice about organization and prioritization. And those choices amplify certain messages.

Your theory would create "a world of lawsuits."

🔥🔥🔥

61/ Stewart: brings up employment decisions?

That totally ignores how the internet works. The whole issue here is how to organize the firehose of info on the internet. Platforms can't scrutinize content piece-by-piece.

62/ Roberts: Makes this exact point -- internet involves millions and billions of decisions. "Every one of those, on your view, brings the possibility of a lawsuit."

Stewart: Too bad for platforms! For Congress.

DUDE -- it was a ~firehose even in the Compuserve days.

63/ Kav: Give me a limiting principle.

Stewart, like Schnapper: Don't worry, there will rarely be an underlying cause of action. ATA is special. Neutral algorithms will protect you in that context.

Kav: You're putting a lot of stock on the liability piece, rather than 230.

64/ Kav: Isn't it better to keep 230 the way it is, and put the burden on Congress? How do we predict the effect of trashing 230?

Stewart: All the courts that reached the 230 consensus are wrong. (!!!)

65/ Stewart: Sorry, Brett, J. Harvie Wilkinson didn't know what he was talking about. Gotta let the heavens fall.

66/ Roberts: I don't buy that, under your 230-trashing theory, we'd only get, like, ATA claims. For instance, wouldn't there now be liability for repeating defamation claims?

Stewart: Naw, I want a ticket for this ride only. "Up next" recs don't raise defamation liability.

67/ Roberts: But the law is not established in the way you're suggesting!

Stewart: Gotta look at Congress's intent. (???)

68/ Alito: Consider a defamatory Yelp-style review. Yelp refuses to take it down. Keeps serving it up. 230 protection?

Stewart: Screw Yelp.

Alito: "You really think that Congress meant to go that far?"

69/ [My last summary of Stewart was a bit mean. Sorry. I think it accurately summarizes where he ended up, though.]

Stewart: Continued attempts to say, look, you're protected for failing to block, and *that's it*.

70/ Stewart now saying that, in figuring out the line between posting and recommending, platforms should get thrown to the wolves of individual states' different defamation laws.

71/ Sotomayor: Naw, GIVE US A LINE. We want a line. Please offer a line.

Stewart: "The laws vary from state to state." 😵‍💫

72/ Why is Stewart fixated on search queries with typos in them?

73/ Stewart: "You'd have to establish the elements of the substantive law."

Well, yeah, that's how it would work -- in a world without Section 230.

The US government wants to break the internet.

74/ Gorsuch raises (f)(4). He likes bringing up AI (ok). Asks when an algorithm counts as "neutral" (valid Q). Does the "let's kick back to the 9th?" thing again.

Stewart: if you do (f)(4)-style sorting functions in way that violates the law, you're out of luck.

75/ Kav: Your key move is to treat algorithmic organizing as recommending. Is that right?

Stewart: What matters is if it's what the user requested.

Once again: THAT'S AN ARBITRARY AND ILLUSORY DISTINCTION.

76/ Barrett: Schnapper tried to distinguish screenshots from thumbnails. That's nuts, right?

Stewart: Yes.

77/ Barrett: What about users and retweets and likes? Same bomb question from earlier.

Stewart: "Those are difficult issues. ... I'm not sure."

FAIL.

78/ Stewart actually makes it worse! Says that retweets are an individualized choice, whereas YouTube "Up next" stuff is systemic.

Dude, you're saying that retweets are LESS protected! What're you doing!?

79/ Stewart to Jackson: Deciding what prominence you give material is not protected.

Uh, ok, but that means this is not about "algorithmic recommendations" or anything like that. You're attacking basic publishing decisions (what's on page A1, what's on page D6).

80/ Lol -- boom! Jackson brings up the subject of "page A1." 😂

Stewart: NOW backtracks into "YouTube is personalized." Retreats to the "super targeted-ness" theory.

Reminder: 230 explicitly encourages the development of the internet.

81/ That's it for the US.

At last: The great Lisa Blatt steps to the podium ...

82/ Blatt: All publishing contains an implicit message. Pets shouldn't be able to get around 230 by pointing to stuff that's inherent to publishing.

83/ Thomas: give me an example of an endorsement that goes beyond 230.

Pretty soft opening Q from Thomas! Good sign.

Blatt: It's a continuum. (Indeed.) Problem is when the platform *adopts* the content as its own. Something more explicit than "Up next."

84/ Roberts: Well, the recommendation is not provided by another content provider.

Blatt: The word "recommendation" is not on YouTube. It's videos from third parties.

Roberts: But the "targeting" is a rec?

Blatt: Nothing in the statute turns on the degree of tailoring.

85/ Blatt: If you go down this road of "Is it targeting?" you fall into the question of "How much?"

86/ Jackson: Wasn't 230, as passed by Congress, narrower than what courts have found, and you're seeking here?

Yikes -- Jackson again *seems* like the *most* anti-230 questioner.

87/ Blatt: 230(c)(1) would do nothing if it was just "publication as an element" or just defamation. This was true even at the time 230 was passed.

Jackson: But what about 230's title? (!!)

(i.e., "protection of private blocking and screening of offensive material")

88/ Blatt: But remember, several of 230's findings go way beyond that title.

(e.g., promote the development of the internet)

And again, really stingy reading of 230 would make 230 do ~nothing.

89/ Blatt: Turns to the bookstore hypo. Proper framing is the bookstore putting horror books in one section, and children's books in another. ✅

90/ Kagan, pushing back: You're still responsible for causing material to get spread to lots and lots of people by highlighting it.

Blatt: But there's no line! All platforms have to choose what to highlight. It's got to be based on something. That's just basic publishing.

91/ Barrett: But what if the sorting is really defamatory? It's really intentional?

Blatt: Yeah, that's protected! [Ooo, refers to 303 Creative.]

Kagan: Does it really go that far.

Blatt: Well, how are you going to carve out "bad" and "good" here?

[Clean, confident answers]

92/ Blatt: You guys know there were lots of algorithmic recommendations back in '96, right? It's just a larger scale now.

93/ Jackson coming in to say that 230 might be about just "good faith" taking down of material. 😬

Perilously close to carrying water for the MAGA reading of 230.

94/ Jackson: 230 doesn't protect the *promotion* of material -- does it?

Blatt: 230 is not coterminous with 230(c)(2)!!

95/ Blatt goes on to cite 230(e)'s exceptions -- which would make no sense unless 230(c)(1) is broad.

Jackson keeps pushing back. Suggests (again I think?) that ALL 230 does is overturn Stratton Oakmont. Just about getting filth off the internet. (Paging Senator Exon ...)

96/ Blatt: Justice Jackson, you're taking an even harder line than the petitioners and the US!

97/ Barrett: What about a discriminatory dating algorithm, as Sotomayor raised?

Blatt: That is not protected. You're *treating* *people* differently based on their status. The harm that would flow is not the speech, it's the conduct. (Basically lays out the Roommates line.)

98/ Blatt: Gotta figure out what the dissemination adds. And "The United States says, 'I don't know, let the states sort it out.'" LoL.

Yeah, not stoked about Henderson (4th Cir.) getting invoked. Blatt: "I agree with like 96% of it. Got a bit lost with the common law stuff."

99/ Alito: softball question (yay?) re: top search results being protected by 230.

Blatt: "Correct. They have to sort content somehow."

100/ Alito: "Would Google collapse and the internet be destroyed" if there's liability for not taking down stuff you "know" to be defamatory?

Blatt: Google'll be fine, but a lot of other sites could go down. CITES YELP. YES. 🚀

101/ Gorsuch raises his (f)(4) question.

Blatt: Yep, love it! But I got a "bit upset when you mentioned the possibility of a remand." LOLLL.

102/ Gorsuch: how do you do a "neutral tools" test?

Blatt: Well, that's not actually the test the 9th Cir. applied. There's an isolated sentence in there.

Gorsuch: But if I'm right, do we need a remand.

Blatt. pause ... "Yeah, I'm gonna have to disagree with that." Priceless.

103/ Roberts going around the bench. Most justices (already) have no more questions. Good sign.

But back to Jackson! ...

104/ Surprise of the day is that Justice Jackson arguably reads Section 230 more narrowly than Justice Thomas.

105/ Jackson: Third-party content placed on *homepage* -- protected?

Well, consider @Wikimedia! That's how Wikipedia's homepage works!

Protected??? Blatt: "Absolutely!"

106/ Jackson: Organizational choices about the homepage can't be isolated and removed from 230?

(Oof!!!)

Blatt: The whole point of 230 is to protect what is inherent to publishing.

107/ And that's it for Ms. Blatt! Well done.

If you judged just by this oral argument (which, to be clear, you shouldn't literally do), this case is heading toward an 8-1 vote with a solo dissent from Justice Jackson. 😳

108/ Schnapper up for rebuttal ...

He just goes, with no interruptions -- a sign the justices are done with him and he's lost his chance to convince them of anything.

109/ And ... the case is submitted!

That's all, folks. Hoo boy! I'm tired.