The very definition of a "gatekeeper service" is that no you're not TRULY free to pick an alternative.

I find that so preposterous. I think you're making it up. Back in 2002-2004 I lived and worked in Italy earning $12k/year as a postdoc. My quality of life there in Italy was substantially higher than when I emigrated to Seattle in 2004 for a $120k/year job. I still live in Seattle and vacation in Europe every other year. My quality of life there is always higher than Seattle.

Let's get away from personal anecdote. Here's a list. https://www.usnews.com/news/be... - these countries are listed as higher quality of life than US: sweden, norway, canada, denmark, finland, switzerland, netherlands, australia, germany, new zealand, belgium, austria, uk, japan, ireland, luxembourg, france, spain, portugal, italy, singapore, poland

The only european countries listed as lower quality of life than the US are greece plus former soviet countries: greece, czechia, croatia, hungary, romania, bulgaria, estonia, slovakia, slovenia, cyprus, lithuania.

Let's look elsewhere. https://worldpopulationreview.... - the story's exactly the same: highest quality of life in Europe, and the US is quite far down.

Maybe there's something really unusual about your personal circumstances that you got a higher quality of life in the US than when you vacation in Europe? It's hard to imagine (1) how that's possible, (2) why you even continue to vacation in Europe.

That remains to be seen. At heart is the question of the "algorithm" by which the publisher decides what to publish. This is similar to subliminal messaging in the sense that viewers aren't aware of the speech they've just heard (going from what you said, that the speech is editorial decision about what to show). But it's different in that with subliminal messaging an objective third party can see what the speech was, but with the "algorithm" no one can see what it was.

Subliminal messaging was ruled NOT to be protected by First Amendment in a Judas Priest case in 1988, but that overruled in 1990. The overruling however (1) sanctioned CBS for failing to disclose the masters, which is closely analogous to ByteDance failing to disclose its algorithms, (2) the overruling was because they didn't prove the subliminal was intentional and didn't prove the subliminal messages were the case of harm, both points on which I think there's clear paths to distinguish TikTok from CBS.

So: I agree it does go to the heart of the First Amendment, but the answer is by no means clear cut.

That's a bad point because there are game-changing differences in what you can do with troll+bot farms vs owning the platform.
1. With troll+bot farms you get analytics based on how users interact with your ads+comments. But if you own the platform then you get maybe a thousand times as many datapoints per user -- these allow for vastly better AI training, say, and better leverage.
2. With troll+bot farms, every post/ad is a concrete thing that everyone can see and trace back to source. But if you own the platform then you can make subtle AND UNDETECTABLE changes to the algorithm, for instance making certain kinds of videos appear 1% more frequently in the feeds of swing voters.

Me: "what happens when milk turns sour?"

Google's first five results: "WHY DOES MILK TURN SOUR? Section 1: Milk is a fascinating beverage that is drunk by every single mammal. Did you know ...? Section 2: we love drinking nice cold milk on a warm day with cookies! Here's more about ... Section 3: milk is produced ... Section 4: You have to be careful about milk going sour; FDA recommends ... Section 5: Why does milk turn sour? ..."

Only in the final section, after pages and pages of inane self-evident drivel, does it get to the actual subject of the page's heading.

1. Timestamp's
2. Full paths to the source code location (which helps a debugger find the source code)
3. Many binary formats have space, eg. Bytes 1,2,3 are used and byte 4 is unused in the header. If the space was allocated by malloc() and not zero-initialized, it will have in it whatever junk was left there before
4. Sometimes if an artifact uses other artifacts built from source, vs artifacts fetched from build-cache since they'd been built previously, it might record that
5. Metadata about the build. I mentioned timestamp. Also host name, env vars etc. they're all helpful for diagnosing/reproducing build issues.

They're all irritating things put in by well-intentioned people for helpful reasons before it was widely understood that determinism is crucial in build systems.

Whenever I want to use the app to interact with an organization, I *never* use the app store. That's because (1) advertising takes top spot, (2) it's hard to distinguish which are the scummy exploitative apps, vs the real ones. Instead I start from the organization's website, search for an "open in app store" link, and follow it, to be sure I'm getting to official proper trustworthy app. This rule change will save me the hassle of indirection.

For instance: search for "seattle center" in the app store to find their app? I get an ad for dating first, followed by irrelevant ones. Search for BA for airline stuff? the app is third place, behind Qatar Airways and Brawl Stars. Search for "seattle metro tickets"? Add for sports betting first, followed by the wrong app, and the correct one is only fourth place.

Getting the wrong app is just an inconvenience, but it's often a 30min inconvenience if I get the wrong app first or have to wade through junk.

A gas plant has been built in 2008. It was uneconomic to run, so they didn't run it, and tore it down, and replaced it with this battery plant. That's the sense in which the gas plant was replaced by the battery plant. I think it's clear and straightforward, and also very well explained in the article. Presumably the battery plant benefits from the existing grid interconnects.

It's right there in the title. A gas plant has been built in 2008 but was uneconomic to run, so they didn't run it. It has been torn down and replaced by the battery system. That's the sense in which the gas plant was replaced by the battery plant. Presumably it was handy to have all the grid connectivity already in place.

You shouldn't care about his opinion because of who he is. You should read and evaluate the strength of his data and arguments. (Unfortunately the link is behind a paywall so I can't...)

For me it's my docked M1 Pro which randomly crashes about once a fortnight. It was hopeless monitor+ethernet on initial third-party docking station so I bought the expensive one listed on the Apple store, CalDigit. This has never been particularly happy with sleep or its external monitor but at least is a bit better. To this day, iMessages on my mac crashes every single time I open one of the chat threads with my kids (which is populated with 100s of screen-time requests which I guess is too hard for iMessage on mac to handle). I honestly had better stability when I developed on Windows.

Q. have you sought to understand? do you have conjectures? what aspects of movies do you imagine appeal to other people that don't appeal to you? You've listed five aspects that appeal to you (pretty, not-being-by-the-numbers, not-being-rehash, novelty-of-visuals, entertaining) but presumably you don't think this is an exhaustive list of aspects that might appeal to others?

It sounds from your quote "I don't understand the level of ongoing interest" that you believe there is ongoing interest (maybe you pick that up from the box office returns) but you haven't yet found yourself able to find reasonable speculations as to why?

The exact opposite.

Aether was an idea about the nature of reality which, as soon as Michelson-Morley got experimental data, was shown false.

Dark energy is experimental data which as yet has no explanation about what it truly is. All we say is "there must exist some underlying reality which will explain these observations, even though we don't yet know what it is".

Clarification: the Tennessee bill explicitly bans *geo-engineering*. It does not mention chemtrails. It was spurred by a government report last year on solar engineering. I think you fell for a clickbait headline from the BBC.

You don't have to be ignorant, backwards, corrupt or inept to be concerned about geo-engineering!

Depending on how you interpret it, it's possible that the United Nations Framework Convention on Climate Change already prohibits geo-engineering. Or possible it mandates it. Hard to tell.

First off, I'd love to pick the algorithm "show only stuff from my friends, in chronological order". I'd hope they'd offer a simplistic no-brainer like this, and it'd be easy to understand.

Second, I'd rely on journalists or academic researchers to study the moderation+recommendation algorithms and I'd go via trusted sources. I'd wait until a review article comes up in Ars Technica and pick there. I think non-technical users would only pick if the question becomes significant enough to filter through to general society discussion. Maybe there'll be some viral videos of TikTok from "influencers" who noticed that they preferred one or another algorithm, and it'll spread through that and word of mouth amongst friends?

Do you have any friends who say "I used to follow instagram but there were too many political articles and I switched to TikTok because it's more fun"? They were switching partly because they discerned a difference in the recommendation algorithm. (which is more or less the secret sauce of TikTok's success).

Third, I wonder why you think end-users are fine at picking up differences between the algorithms used for internet-search (google, bing, altavista), and for map-routing (waze, google maps, apple maps)? How are they basing their choices? Why won't the same kind of end-user choice about subtle and complicated algorithms also apply to feeds?