Big Tech's future is up to a Supreme Court that doesn't understand it
axios.com
axios.com
What "big tech" is doing is not new it's the same game that companies have been doing since there have been companies. Common carrier status was always about not being party to messages of your users. Big tech is absolutely party to the messages of their users, even private ones. We all know it, they do it intentionally and even the stodgy old people on the supreme court can clearly see their game.
If you're going to "control the narrative" by choosing what people see and what the don't, then you need to be responsible for what is displayed. This isn't and opt-in level of control there's not even an opt-out. They blatantly decide what the news is and feed it to you, no matter what you try and do to search around it.
It has always blown my mind that google maps has no option to show all the points of interest no matter how far you zoom in. Just one tiny example of a service populated by users but then controlled by the machine.
Be a common carrier or don't be. This manipulation game needs to end. I don't know what the court will do but hopefully provide guidelines on what common carrier is.
You would actually prefer to get on Youtube and see the thousands of awful videos uploaded before the one that’s related to what you’ve recently watched?
From my point of view, I think that losing it would be a net win for society as a whole.
I use youtube -daily. But I see it for what it is.
Libraries seem to function just fine and people find things with indexing but without force feeding a narrative.
That's how YouTube started and it worked great. Now I have to fight the site to find content because it's been demonetized and YouTube won't suggest it anymore...
As a technical person, I am actually sympathetic to the plaintiff's argument that Section 230 may not necessarily protect recommendation systems designed for engagement maximization, and particularly not in the case of push notifications or the like.
> If they lose the case it won't be because the Supreme Court doesn't "understand".
The honest truth is that if they lose this case it will be because Big Tech pissed off conservatives. The court is in the midst of a legitimacy crisis. If the courts don't pull out of the legitimacy tail-spin and return to a situation where at least issues of commerce are less politicized, the US may start facing some problems reminiscent of emerging markets with weaker rule of law/capital protections.
The first is whether section 230 is being correctly interpreted. This is undeniably the domain of the Supreme Court. In my opinion, section 230 is being correctly interpreted; moderation does not somehow remove the immunity. This bit seems pretty salient on that point:
> No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.
The second issue is whether section 230, as it exists, is still a good idea. That question _should_ be for Congress to decide, not the Supreme Court. I'm also sympathetic to the idea that at some point, moderation can start to look a lot like publishing and perhaps should be treated similarly. The current law seems pretty clear that that's not the case, and a ruling that changes that starts to look an awful lot like writing laws by strange interpretation.
It's all kind of paired with the inability of Congress to get anything meaningful. People have gotten used to Congress being ineffective, and are now looking at the SC to effect change. It's a dangerous line to tread, though, since federal judges are appointed rather than elected and don't have a direct feedback mechanism for people to express discontent.
This is a distraction and non-issue... “The court might think it's doing one thing and it's actually doing something very different,” said Evelyn Douek, a law professor at Stanford who specializes in tech law. “It’s ill-matched to the problem.”
If your algorithm can't be explained to a judge, someone who is highly educated at understanding logic and reason, then perhaps it is broken anyway.
The fact is Big Tech has dug its own grave long ago, and that should have been obvious since the Cambridge Analytica scandal.
> If your algorithm can't be explained to a judge, someone who is highly educated at understanding logic and reason
This might be true of appointed judges. It certainly is not true of elected ones.
Certain types of promotion and downvoting are clearly within the legislative intent of Section 230. Suggestion may not be, and I am sympathetic to the plaintiff's argument that recommendations -- and particularly push notifications and the like -- may not be covered by 230.
(As always, it would be much better if the legislature simply did their damn job, but we seem to be content with permanent gridlock.)
> If your algorithm can't be explained to a judge, someone who is highly educated at understanding logic and reason, then perhaps it is broken anyway.
I have no doubt that many in the legal field have this opinion -- particularly judges -- but you seem to be wildly under-estimating the degree of innumeracy on the bench. And, indeed, through most of the America's professional class. Most of the lawyers that I know haven't taken a mathematics course since high school, and one has never even taken a Calculus course (his college mathematics requirement was essentially a repeat of high school algebra). Neither of the judges I know are capable of even non-trivial computer use. Technological illiteracy really is an apt metaphor.
Within the legal community, somehow, wanton ignorance of the very basics of human knowledge in various fields is considered a major deficit. But technological illiteracy is if anything a badge of pride and a sign of status.
I have no doubt that these people are capable of understanding the technical details, but I'm also highly skeptical that they have the time for such an enormous undertaking given the typical state of numeracy and technical literacy on the bench.
Does this mean they need to understand Calculus? Not at all. They need to understand the algorithm in a way that pertains to the law. Not in a way that they could rewrite it. This is a common misconception among those in the tech industry. Nobody really cares about the minutia like you or I would. What they care about is the end result. so...
Does the "suggestion algorithm" violate section 230 or not? The ones I have interacted with and researched I believe do. But the lawyers don't have to convince me. They need to convince a Judge. A person designated by the republic to uphold the law.
This is true, but one is federal. And substantially younger than most of the SCOTUS justices. He doesn't know how to use a web browser. Which, to be honest, was actually unbelievable to me, since even the older clients at the homeless shelter I volunteer with know how to use a web browser. How do you even function in today's world without this knowledge? I guess a job for life and a rotating staff of ambitious brown nosers who are willing to google things for you is the answer.
Of course judges don't need to be experts on everything. But you seem to misunderstand my critique of the bench. This isn't akin to "can't write a novel" or "never read Shakespeare"; it's more akin to "actually functionally illiterate in anything outside of law and some subset of the humanities".
You might not care that a judge hasn't read this or that book, or never wrote a short story, for example. But if your judge didn't know how to read?! You might have some serious concerns. That's the state of affairs with respect to numeracy and technology.
For that reason, the rest of your post seems like fairly non-responsive to my primary concern.
In particular, I don't doubt the raw capability or intelligence of SCOTUS justices (or other judges). Becoming a federal judge is incredibly difficult; these are bright and highly motivated people.
But I do think many judges lack the basic numeracy and technical literacy required to know what they don't know, or learn what they need to learn. Again, the baseline here is less like "never written code" and like "doesn't know how to use Chrome".
A secondary concern is a lack of humility, which is of course a precondition for learning. Combined with a sort of general innumeracy among the elite set. I can give some examples of opinions that put this combination on display.
Yup! I'm proud to say the algorithms we designed in the healthcare space are 100% explainable and deterministic to staff, executives, and lawyers. Everyone that uses our software knows exactly what went into the decisions we made, and they can always make the determination for themselves using all the same data and process.
When hiring, we made a point about deriding FAANG for their stupidity* of throwing their hands up, saying healthcare is too hard, and letting AI do it. It also had a way of filtering out careerists, and a way of finding people with very deep respect for data privacy and security. :)
* Yes, stupidity. Haters can fuck off. There's no other valid excuse you can give that won't make a judge say the same, albeit more politely. https://www.youtube.com/watch?v=ww47bR86wSc
>Issue: Whether Section 230(c)(1) of the Communications Decency Act immunizes interactive computer services when they make targeted recommendations of information provided by another information content provider, or only limits the liability of interactive computer services when they engage in traditional editorial functions (such as deciding whether to display or withdraw) with regard to such information.
>Issues: (1) Whether a defendant that provides generic, widely available services to all its numerous users and “regularly” works to detect and prevent terrorists from using those services “knowingly” provided substantial assistance under 18 U.S.C. § 2333 merely because it allegedly could have taken more “meaningful” or “aggressive” action to prevent such use; and (2) whether a defendant whose generic, widely available services were not used in connection with the specific “act of international terrorism” that injured the plaintiff may be liable for aiding and abetting under Section 2333.
[1] https://www.scotusblog.com/case-files/cases/gonzalez-v-googl...
[2] https://www.scotusblog.com/case-files/cases/twitter-inc-v-ta...