Replying to @null and @amyvdh
Thank you Amy, big hugs and happy new year!
January 2023
253 tweets
Replying to @montezumachavez and @alextcone, @Chronotope
I am proud to know all of you, and Luis we need to all meet you this year!
♥ 3
Replying to @TzviyaSiegman
Thank you ♥️ I hope you have an excellent year and that we get to hang out.
Replying to @cyberbaloo_
Toi aussi, gros bisous!
Replying to @michael_nielsen and @kate_sills
An essential aspect of the good life is that it is skillful. A list of books, skillfully assembled, can be a great source of self-cultivation and I would expect that from a (good) syllabus. I think that's different from "let's power through a list of famous books."
♥ 3
Replying to @BuggeErik
We have to! Happy new year! I'm glad we got to talk a few times in 2022, it's not as great as meeting but better than nothing!
♥ 1
"Instead, we allowed corporations to set their own priorities. They pursued wealth and power, with strategies that aggravated every problem in society. No industry did more harm than tech."
Quoting a tweet by @Moonalice ↗
♥ 7
Replying to @Moonalice and @EvanSelinger, @latimes, @FoxCahn
I would love to hear your thoughts too, figuring this out is very much in my area of interest!
♥ 3
"Gatekeepers may no longer control what gets published, but algorithms control what gets circulated."
Beautifully succinct description of the problem we have with the absence of pluralism in algorithmic media (search & social). I really look forward to reading @wrbucla's book!
Quoting a tweet by @nils_gilman ↗
♥ 23↺ 13
Replying to @chrismessina and @danbri
I asked Dan because he is the Guardian of Web History, and because he's both a good friend and a curious historian, he went to ask you :) I didn't know you'd worked there.
I'm thinking about what browsers could/should be if we want the Web to succeed.
♥ 2
One thing that I believe they need to do is to commoditise search and social — or be commoditised by them, which is what's been happening and isn't a promising direction.
Flock is one obvious place to look for prior thinking :) Thanks a lot for the shot, exactly what I needed.
♥ 2
Replying to @danbri and @chrismessina
Only nerds ever used folders, Dan, everyone else just dumped crap on the desktop and occasionally rolled it up in an "OLD STUFF" directory.
Browser UI is bad, but I'm even more worried about how in-app social/search is eating it. It's not done, too.
♥ 1
Replying to @chrismessina and @danbri
This is super interesting, thanks! Almost fifteen years later it still feels relevant. Identity in browsers today ranges from meh (Firefox) to actively user-hostile (Chrome). There's a lot of room for something better.
Doux relents de peste brune chez @Cdesarts où l'on déplore le "wokisme" et ses "arguments fallacieux de justice sociale et d'égalité raciale" (ah...) tout simplement parce qu'une expo a osé placer ses pièces dans leur contexte.
Encore un qui n'est pas entré dans l'Histoire...
Replying to @ThatAndromeda and @danbri, @chrismessina
I mean, we could talk about @swodinsky's document...
♥ 2
Replying to @chrismessina and @danbri, @ArcInternet
I agree it's super promising; I'm still withholding judgment because it's still early and has some clunky, but I like what I've seen so far.
Replying to @ka_iwanska
It's still not high enough. Those companies are operating with supracompetitive margins that are easily within or above that range — and breaking the law helps support a monopoly position (because who else can feed that many lawyer?).
♥ 3
Say you're even "only" getting 20% margin. It takes multiple years for a decision to happen but the fine only takes one year's turnover, so you're still profitable.
And *even* if all the profit were taken, you've still used the leverage to kill the market, you *still* win.
♥ 3
You want to make this actually work? We need to talk about making the people directly responsible — at least financially. I don't believe in prison, but I won't complain if Zuck & Sundar go to jail, they deserve it more than most of the people there.
♥ 2
Replying to @gateklons and @ka_iwanska
I see no reason why it would only hit Irish execs?
Clawing back the money would be nice, but I think that it would be difficult. It also doesn't bring back the law-abiding competitors you lost!
♥ 1
Replying to @gateklons and @ka_iwanska
I'm not a lawyer but I would assume that if you're a senior accountant at Meta Ireland and there's an email from Zuck saying "do fraud!" then he'd be on the hook as well.
♥ 1
"The Expanding Dark Forest and Generative AI" in which @Mappletons makes a compelling case that LLMs et al. will strengthen the dark forest web.
maggieappleton.com/ai-dark-forest
♥ 8↺ 3
One interesting consequence is that behemoth one-size-fits-all will keep losing relevance as a result. We should be building social & search specifically for a dark forest web.
♥ 1
I have to say that this Finding Out phase of the Let's Fuck Around Making Friends With Fascists project that Republicans have been running isn't entirely unpleasant to watch.
♥ 11
Replying to @riptari
I'm not sure that's 100% fair, she's building on her previous maggieappleton.com/cozy-web which is very much about that.
♥ 1
Replying to @null and @icopilots
I guess?
Replying to @le_hasard
Je crois que ça se passe au niveau du système de fichiers…
Replying to @le_hasard
Ah ouais mais tu fais des trucs radicaux, aussi.
Very excited to see how this research programme into democratic microfoundations develops (via @mariafarrell). It feels badly needed.
"Analytical Democratic Theory: A Microfoundational Approach" by @henryfarrell, @hugoreasoning, and Melissa Schwartzberg
cambridge.org/core/journals/…
♥ 6↺ 3
@RollingStone brings together a stunning investigation into war crimes in Bosnia with a state-of-the-art photographic provenance and archival system to help ensure that these events remain authentically documented.
Quoting a tweet by @RollingStone ↗
It's time to start this year's book thread - welcome to 2023! ♥️
♥ 16↺ 3
First, links to the previous threads.
2022
♥ 1
2021
♥ 1
Read the whole thread: 32 tweets, continued until 30 Apr 2023 →
Replying to @montezumachavez and @Chronotope, @alextcone
No one knows more about this than you, Luis!
♥ 5
Replying to @kkomaitis
For those who want to follow the discussion, it's over at the cool place! mastodon.social/@robin/1096651…
♥ 1
Given how much of my social life consists in interacting with people who are either on a screen or on a sheet of paper and given that I understand how to use a mask, etc. you might *think* that I wouldn't be able to get covid a second time if I tried.
Well, you'd be wrong...
♥ 16
Replying to @kaveh1000
Thank you!
Replying to @null and @amyvdh
Thank you! It seems to be better than last time.
Replying to @megangrA
Thanks Megan! So far it's less bad than the first one so fingers crossed I won't need anything other than rest and soup.
Replying to @clancynewyork
Thanks! Yes! I think it must have been at the supermarket, where I was wearing a KF94. No other recent real social contact other than the kids and they are negative. So far I'm mostly tired with a headache.
♥ 2
Replying to @alexanderdanilo
Given the headaches I think I'll use that medication for the after party!
♥ 1
Replying to @mberk
Thank you, internet hugs back!
Replying to @wheresaddie
I want to blame the kids, you know, but they just keep testing negative. Maybe these are Immaculate Infections and we should start a miracle souvenir shop? Or maybe we're in one of these bad plots in which book characters are somehow alive and we're getting it from them?
♥ 1
Replying to @dietrich and @BrendanEich, @boscolochris, @darrello, @csuwildcat, @gordonbrander, @brave, @shivan_kaul, @fmarier, @lidelOrg
I think that the policy we come up with has to depend on which promises we make in the ipfs: scheme. The SOP is needed because the HTTP web is full of side effects and so we clumsily restrict interaction to a smaller, safer group of side effects.
♥ 4
If we can guarantee that ipfs: loading is nicely composable (I'm handwaving on details) then I don't think that we need an SOP. Just load IPFS in IPFS all you want. Things are less clear with ipns: and we need to think carefully about IPVM interactions (and HTTP loading).
I'd love to hash the details out! This would benefit from some security/maths chops too.
♥ 1
Replying to @Chronotope
Thanks, it's OK, definitely less bad than the first time!
Replying to @BrendanEich and @dietrich, @boscolochris, @darrello, @csuwildcat, @gordonbrander, @brave, @shivan_kaul, @fmarier, @lidelOrg
Just to make sure we're on the same page: you're considering eg. the <script src="ipfs://deadbeef/jquery.js"> case, right? That wouldn't hit an SOP wall over HTTP so I don't think that it needs to be treated differently over IPFS?
♥ 2
We need to specify the cases in which it should/shouldn't trigger mixed content issues. And we need an in-depth look at the privacy properties of such a load — the notion of "third-party" is pretty meaningless on IPFS!
♥ 2
Replying to @null and @fran_mady
Merci!
Replying to @null and @MadronaCitizen
Yeah, it's definitely milder. Given how beaten up I was the first time, boosted and all, I'm pretty sure I wouldn't have made it through unvaxed.
Replying to @lidelOrg and @dietrich, @BrendanEich, @boscolochris, @darrello, @csuwildcat, @gordonbrander, @brave, @shivan_kaul, @fmarier
I think we need to be a bit more specific, unfortunately! If you're looking only at eg. script@src=ipfs:cid then yes. But if you consider <iframe src="ipfs:cid/?foo=x"> then you have the problem that this could transitively load from HTML, pass the param, and mutate remote state!
♥ 1
That's what I mean by hashing the details out: there are IMHO very cool promises that we can make with IPFS in terms of changing the SOP but we need to fully work stuff out. Eg. iframed IPFS should load with CSP "default-src 'self' ipfs: ipns:; img-src 'self' ipfs: ipns: blob:"
♥ 2
Replying to @lidelOrg and @BrendanEich, @dietrich, @boscolochris, @darrello, @csuwildcat, @gordonbrander, @brave, @shivan_kaul, @fmarier
Absolutely, but what would normally fall under "third-party" threats is the ability to identify (probabilistically enough) the person who triggered a load so as to track their behaviour. So the threat is the same for all IPFS loads you don't already have.
♥ 1
That's different from HTTP where the threat will differ based on which domain is top-level.
For mixed content, so long as you verify what you load (and there are no credentials over IPFS), then loading ipfs: in an https: context should always be safe?
♥ 1
And same with ipns: assuming that you verify that it was signed correctly too. I could be wrong (I haven't looked through the full mixed content threat model) but this is pretty neat.
♥ 1
Willow turned out to be much better than I expected (particularly characters and dialogue). Boorman 💖
♥ 2
Replying to @BrendanEich and @lidelOrg, @dietrich, @boscolochris, @darrello, @csuwildcat, @gordonbrander, @brave, @shivan_kaul, @fmarier
If you're sourcing the script into a context that can have side effects, it's not safer any which way. That's why I keep returning to the HTTP/IPFS boundaries. The risk profile when crossing between them isn't the same as when encapsulating in one (even if embedded).
♥ 1
Replying to @julien51 and @netflix
Julien, our current governance systems are the root of lots and lots of pain and evil.
The Topics API (from the "Privacy Sandbox" set of proposals) has been reviewed by @rhiaro on behalf of @w3ctag. There's… not much left of it standing.
github.com/w3ctag/design-…
♥ 12↺ 6
Has anyone run some pre/post-Elmo takeover engagement numbers on their tweets? It's anecdata, but things feel pretty off.
♥ 4
Replying to @alextcone and @rhiaro, @w3ctag
I don't think it's user agent principles that keep anyone from participating. I think that the case that this has actual value that would justify the trade-offs hasn't been made. It remains unclear to me that we need cross-context topic targeting.
♥ 5↺ 1
I don't see it helping quality publishers. If it is effective enough to help service publishers who can't build audiences easily (eg. TimeAndDate) then it also helps fund disinfo. If it doesn't help disinfo, then it doesn't help those either.
♥ 2
Replying to @alextcone and @rhiaro, @w3ctag
If it's going to happen even without a Topics API, then we should *definitely not* add the API as it just makes things worse with no upside.
♥ 4
Replying to @mberk
Right, same. And by engagement I mostly mean people replying and the such. And now there's nothing. Mastodon has great convos happening but not everyone is there and I'm missing quite a few folks!
♥ 2
Replying to @alextcone and @rhiaro, @w3ctag
Sure, I don't disagree, but that only works if we're actually controlling the space!
♥ 3
Replying to @mberk
I think I have rebuilt about a third of my network, mostly via using I forget which of the tools that finds Mastodon links in people's bios. I did a mass refollow. But still missing some key folks and it feels!
♥ 2
Replying to @thazel_thazel
Haha 😂
Replying to @sangwhanmoon and @rhiaro, @w3ctag
That's what I assumed and I think it's also what I said?
♥ 1
Replying to @sangwhanmoon and @rhiaro, @w3ctag
Ah, that's not how I read that formulation, to me it's "drafted by X but with the authority of Y." But I see your point, and I certainly didn't mean to imply that this was Amy gone rogue, only to credit their clear and level-headed drafting.
♥ 5
And thanks for clarifying that this was an even broader discussion than usual reviews (which seems quite logical).
♥ 3
Replying to @mberk
Thanks! I have a few things I'm looking into, it's been 20 years since I read Aristotle so I'll see about that too. My concern is that he has a tendency to say something amazing and then go: "and that's why women, barbarians, and slaves are naturally inferior."
♥ 2
As an adult, I can read past that to some extent, but I know that that's all that my kids would see, even with a warning I think.
♥ 2
“We believe that improved privacy should ultimately lead to better performance,” McLaughlin said, “whether that’s brand lift or more direct results.”
adexchanger.com/privacy/this-v…
♥ 3
"And there is already early evidence that prioritizing privacy can be good for performance. Companies with responsible data practices typically see a 23% increase in purchase intent, according to research released last year by Magna and data governance platform Ketch."
♥ 1
I don't know if their methodology is any good (measuring this from the outside is notoriously hard) but this is an interesting development. Adtech is changing.
♥ 2
Me planning to write: Haha fuck yeah!!! Yes!!
Me writing: Well this fucking sucks. What the fuck.
♥ 48↺ 2
Replying to @__apf__
Go bold, then it actually looks good that it's growing out of your head. The NYT's HR head had an amazing colourful painting right behind her, and it really worked.
♥ 2
Replying to @null and @MadronaCitizen
As slow, more painful.
The Internet is allowing us to build a richer, more complex society but the way in which we Internet today is failing to support the governance systems that a more complex world requires. I take a look at why and what's ahead:
berjon.com/internet-trans…
♥ 95↺ 23
Because fuck yeah it's Friday evening and you wanted nothing more than to spend your weekend reading me ramble out 5,000 words!
♥ 15↺ 2
Replying to @hibaymj
I've promised another ~7k over the weekend so I had to cut it short ;-) If you look towards the end, there's a paragraph or two that could easily double the thing if I did them justice, but this has been in the pipeline since ~June, you have to stop somewhere!
♥ 3
Replying to @mberk
For me the worst is putting something coherent together. Material is never a problem, but turning it into something coherent and interesting, and not making every paragraph a research project that reinvents metaphysics from the ground up is the worst.
Totally the kind of experience you give when you care about privacy, right?
♥ 37↺ 11
Replying to @megangrA and @economistivers
I'm still too in covid to have the strength to dig into what they're doing, but the linked privacy policy had stuff about trusted partners and sharing data.
♥ 1
Replying to @null and @siobhansolberg
YOLO!
Replying to @vkw and @larakiara
Victor, I'm all for productive disagreement but that research is just lobbyist fodder. Citing it really makes it look like you haven't done your homework on what is a rather important topic. Maybe talking to actual publishers would help?
♥ 10
Chrome 2020: Unlike other browsers we could talk about, we're going to do our privacy work through a standards process.
Chrome 2023: Well, I mean, obviously the "whenever the standards process agrees with us" was implied there.
Quoting a tweet by @larakiara ↗
♥ 48↺ 25
I guess that maintaining the revenue stream from disinformation - which is a key part of what Topics does - is more important to Google in a downturn than keeping the standards process credible.
♥ 18↺ 6
Replying to @null and @MadronaCitizen
I know, but as much as I love CSS, it matters a bit less than structuring the economics of the Internet, including bits that are vital for democracy.
♥ 3
Also, "it wasn't done through standards" was their excuse for not doing tracking protection. You'd think they'd at least have the honesty to acknowledge the hypocrisy.
♥ 1
Replying to @null and @MadronaCitizen
You're not wrong. There are exceptions but the typical vibe is definitely that.
♥ 1
Replying to @null and @MadronaCitizen
The worst I've seen was in the AMP stuff. This poor guy was thrown into a group where he had to fight sensible proposals just to keep their crap alive. He just read lawyer statements, you could tell he hated it.
Replying to @null and @MadronaCitizen
Yeah I totally felt for the guy. They probably told him that working on AMP would be fun and important (think of the scale!). I absolutely give him credit for not buying it.
Replying to @null and @MadronaCitizen
Right, that makes sense, but it's not something where you'd stubbornly support a bad idea forever.
Replying to @Log3overLog2 and @vkw, @larakiara
This is an ad industry selection of papers most of which answer the wrong question (comparing cookies vs non-cookies in a cookie world instead of a cookie world vs a non-cookie world). Like, maybe you have valid reasons but let's not pretend that this is one.
♥ 5
Replying to @vkw and @larakiara
The subject is Topics but you keep making arguments about all of cookies or all of targeting. That's just besides the point?
Online gamers are a high value segment. In what way did empowering ad networks to target your gamers on disinfo sites help you?
♥ 2
Replying to @Log3overLog2 and @vkw, @larakiara
We would all love it, but we don't have it. Replacing it with data that we have no reason to believe provides a good proxy doesn't lead to better decisions...
Methodologically, the closest attempt is Acquisti. I'm still not 100% convinced but the difference is very notable.
♥ 3
Replying to @Myles_Younger and @dmarti
Totally, I'm not saying that there's nothing good in their proposals - there is. But they only use standards to pretend (possibly to themselves) they don't own the web. I'm tired of that joke.
♥ 4
Replying to @Log3overLog2 and @vkw, @larakiara
You're right, I was misremembering this as the study that sought to separate out the specific things that cookies do. It is indeed like the Google study: a high powered answer to an irrelevant question! I'm afraid this brings my confidence in that slide even further down 😂
♥ 1
Replying to @alextcone
Ah, is it doing that thing where the tab reloads when unfocused and doesn't return to your place? I *hate* that. Firefox fixed that thankfully, it used to drive me bonkers.
♥ 2
Replying to @Log3overLog2 and @vkw, @larakiara
I beg to differ. To begin with, they're designed as one sided: they don't (and can't) measure the loss that the system creates. Then they don't break out the value of different cookie functions. Maybe there is value but all of it is from, say, attribution?
♥ 5↺ 1
They make a big bag that they call "publishers" but offer no useful details about variability (which is, as far as I can gather, high). Finally, there are way too many factors that they don't control.
So the number could be different to the point of changing the sign.
♥ 3
A fix that addresses just a fraction of the use cases could be sufficient. Or a fix that is specific to one industry or product class could be sufficient. Or there may just be too much noise.
It's irrelevant to the question of Topics because it doesn't inform a decision.
♥ 2
So "we're going to ignore the consensus of the web community because we have these dodgy numbers" doesn't strike me as a great example of responsible stewardship of the platform, sorry.
♥ 2
Replying to @alextcone and @Log3overLog2, @vkw, @larakiara, @pbannist
Well, take a random sample of the sites that benefit from these and visit them. I'm not saying that it's *all* disinformation, but...
♥ 1
Replying to @alextcone and @Log3overLog2, @vkw, @larakiara, @pbannist, @AppNexus
Right - I know you know what's in there!
♥ 2
Replying to @alextcone and @Log3overLog2, @vkw, @larakiara, @pbannist, @AppNexus
I didn't say it was conspiracy, I'm just pointing out that Google makes brisk business funding disinformation and it is predicated on precisely this. I also have too much respect for the intelligence of the people involved there to not assume they know this is a bad solution.
♥ 2
But since it's Saturday night, let's entertain some conspiracy, or at least some kremlinology!
I think that Google oversold to the CMA their ability to replace everything the cookie does, and included stuff that could legit die in the mix, like Topics.
♥ 2
With everyone (predictably) pointing out flaws in Topics, they had a choice to:
A) Drop the Sandbox.
B) Betray their commitments to the community and to privacy.
C) Betray their commitments to the CMA.
They chose B.
♥ 2
Now, this is 200% speculation and it's arguable that given such a shit choice, B was the least sucky option.
Still, there were paths that didn't lead to this trap.
♥ 2
Replying to @pbannist and @alextcone, @Log3overLog2, @vkw, @larakiara
Precisely. Which is why I don't think that this mechanism should just allow funds to flow from good to bad. It would be better to map ways to wall off what behaviour sharing, if any, happens.
♥ 2
Replying to @pbannist and @alextcone, @Log3overLog2, @vkw, @larakiara
I know, right? It's almost like someone there thinks that it's not good for publishers.
♥ 3
Replying to @Log3overLog2 and @pbannist, @alextcone, @vkw, @larakiara
I thought they stopped using that enforcement method in 1672?
♥ 2
Replying to @Log3overLog2 and @pbannist, @alextcone, @vkw, @larakiara
I had you down as more curiosity driven.
♥ 3
Replying to @Log3overLog2 and @vkw, @larakiara
I'm not claiming any mantle, I'm just pointing out that there's clear consensus that Topics aren't a good idea.
♥ 2
Replying to @Log3overLog2 and @vkw, @larakiara
The whole point of having a standards community and process in order to make decisions by consensus is precisely to avoid having one guy who thinks he alone has found the right balance calling the shots for everyone. But, and I mean this, I appreciate the honesty.
♥ 3
I certainly have a lot more respect for what you just said than for all the lip service about caring about standards that seems to only really matter when it's Apple doing something.
♥ 2
Replying to @Log3overLog2 and @alextcone
I have tweets I wrote in the past couple of weeks that I can find because they have notifications attached to them, but they don't show up in the timeline on my profile. And digging up tweets from more than a few months ago is a crapshoot.
♥ 2
I also got no mobile notification for either of these replies, I just saw them in the notifications tab by chance.
♥ 2
🎶 🎶
Crème
Brûlée
Au vent
Des landes de p-
♥ 4
Replying to @odemeulder
VOILÀ!
Replying to @null and @0xcccrrr, @bmann, @blaine
HA! Can't tell you how much I love this comment :)
♥ 1
Replying to @null and @0xcccrrr, @bmann, @blaine
Working on a book involving scientific forestry 😁
♥ 1
Replying to @mberk
Quoting @LAM_Barrett, "notice and choice, a method of privacy regulation which promises transparency and agency but delivers neither." If I recall correctly in the same paper she says "it's not notice and choice, it's take it or leave it."
♥ 1
And it really shows everywhere. It's a system of structural learned helplessness: you are continuously made to know, and made to know that you can change nothing about it.
♥ 1
Replying to @blaine and @0xcccrrr, @bmann
I think that's unavoidable! 😁
Replying to @montezumachavez and @megangrA, @Chronotope
I wouldn't read removing editors as withdrawing resources, there were arguably more than needed. Also, this is pretty mature, it's mostly lawyers who need to catch up 😂
♥ 2
Replying to @Log3overLog2
Thank you! This does indeed look very enticing, and anything like A Memory can only be good. Hahaha, so much for my personal moratorium on book buying. It's funny how this piece seems to bring out reading recs (and interesting ones too).
♥ 3
If you didn't catch it over the weekend, start your week off right with a little read on the Internet and planetary governance!
♥ 4↺ 1
Replying to @KrisShrishak and @clancynewyork, @ICCLtweet, @Europarl_EN
Actually, I think that this hasn't been the case for a few years now! Ever since they made native replies.
Replying to @clancynewyork and @KrisShrishak, @ICCLtweet, @Europarl_EN, @luca
Huh, fascinating and weird. Thanks!
Replying to @danielzarick and @can
Same. Addition is trivial on a rounded tip; calculating 20% exactly is work.
Replying to @LMSacasas
There is no other opener. Also, I love subtle differences in quality. The screw can vary greatly in how smoothly it enters cork, and the last little articulation sometimes doesn't bend the exact right amount to catch on the rim properly.
♥ 2
Kind of surprised you didn't label this "Tool For Conviviality" though!
♥ 1
Replying to @ehafen and @RepublikMagazin, @Adrian_Perrig, @ruddyconsult
Thank you!
Interestingly different take (compared to berjon.com/internet-trans…) on planetary governance. RMU seems to only consider either world government or a society of nation-states. No transnational civil society for commons, etc.
Quoting a tweet by @NoemaMag ↗
♥ 1↺ 1
Replying to @null and @benharnett
That's what you get for wearing shoes inside the house.
Replying to @NBerggruen and @mangabeiraunger, @NoemaMag, @nils_gilman, @JonathanSBlake
At the risk of looking like a shameless plug, I happen to have published berjon.com/internet-trans… on this topic just last Friday. That world government is a bad idea seems pretty clear to me, but I am less convinced that operating only between nation-states is the one alternative.
There is space for nonstate transnational topic-specific governance. Such institutions exist, and not necessarily thanks to voluntarily delegated power by nation-states. All are imperfect, but several are reforming themselves.
♥ 2
One interesting aspect is that these can then cooperate with governmental systems at various scales. Eg. on digital accessibility, maybe the nation-state isn't interested, but a regional education system could be. Subsidiarity adds up in potentially powerful ways.
I work in internet governance, sit on the W3C board, etc. so evidently I'm horrendously biased and informed by that experience. Still, I think a richer framing is possible!
♥ 1
Replying to @nils_gilman and @NBerggruen, @mangabeiraunger, @NoemaMag, @JonathanSBlake
I look forward to reading this! I don't seem to see information about it, I assume it's too early?
Reminder as the #OnlineSafetyBill seems headed forward that it's possible to support child-safe processing without age verification. berjon.com/gpc-child-priv…
♥ 13↺ 8
Replying to @othermaciej
That's a very important question. I think that the distinction is that we should use GPC (or equivalent) for everything that has to do with broadly technical safety (privacy, but also things like profiling, types of recommendation systems) but must absolutely not do so for...
♥ 1
...content or anything like that. Otherwise you disincentivise use in adults and lose the hiding in the crowd effect. For porn, etc. you would still need some age gating (if you accept the premise of age gating in the first place of course).
♥ 1
Replying to @PryvitKyle
In practice, I think no, I think that kids and adults should have the same and it should be the default, but shhhh 😉
I'm not sure that adults can consent to most of this either tbh...
♥ 3
Replying to @LMSacasas
There's a dynamic on top of that dynamic. Right now they embed a watermark in their images to avoid at least learning directly from their output. But people have started using the same watermark in human creations.
♥ 1
Replying to @null and @MadronaCitizen
Absolutely, and their ability to change actors for the same character was really interesting. But both got cancelled — it looks like there isn't a business model for scifi unless it's the nth iteration on an old franchise :(
Replying to @null and @MadronaCitizen
True, but they did cite costs for Altered Carbon. And The Expanse got cancelled by two different networks, one of which has a shit ton of money. And I can't think of non-franchise scifi (at least effects-heavy scifi) that made it?
Replying to @null and @MadronaCitizen
Yeah, that's been weird since they announced it. I think we should just call it cancelled — maybe it'll make a comeback like Willow, but until then…
Replying to @faineg and @_Jordan_Green
Eating by myself at the hotel bar is the best part of any business trip to any country. Read a few pages, have a little chat with the bartender, wrap up early and go read in bed — just the best evening.
I was never clear on where this fear thing came from.
♥ 2
Replying to @null and @benharnett
That's amazing.
This thread is so intense in train wrecks. Chat AIs would be the most boring overhyped non-event since boo dot com were it not that people insist on putting them to use in the dumbest, more dangerous ways.
Quoting a tweet by @ZaneGTCooper ↗
♥ 11↺ 6
Replying to @null and @benharnett
The capybara tho.
Replying to @riptari
This is beautiful. Amazing tears. I hope there's more where that came from.
♥ 1
Is there a machine-readable rights reservation protocol to enforceably exclude derivative works by LLMs et al.?
♥ 9
23 years later, the battle rages on.
♥ 9
Replying to @espadrine
I know about the watermark, I've heard people are using it to protect their own works. I'm interested in the legal part because I think that the old search deal of content for traffic is no longer fair and equitable, and this makes it worse.
♥ 1
Replying to @mikarv
Right. I'm fine with research, I just don't want my work to be used commercially for what I consider to be unacceptable purposes and with no value in it for me.
♥ 2
Replying to @MissIG_Geek
Sure, but flash backs from other planets, people travel there, Martian artefacts, etc. Thanks, I will!
♥ 1
Replying to @smartopian
I hear you but nothing prevents a browser from making it the default. And then it's bas the same!
Replying to @lmrlaurent and @mikarv
I absolutely do, but last I checked it didn't support specifying purposes and I think that's still the case, no?
More stupid in a long line of stupid from the stupid factory. I wonder if some of those clients will switch to Mastodon?
Quoting a tweet by @kottke ↗
♥ 1↺ 1
Replying to @MegBastard and @MelissaSweetDr, @JenniferDoggett, @blinkandumissme
Thank you!
Replying to @remitbri
Nice!
Replying to @sara_hjelm
There's nothing selfish about wanting hope! There's work to do but we can make it happen!
♥ 1
Go fuck yourself, @TwitterSupport.
♥ 4
Replying to @ForEducation5 and @TwitterSupport
And stop asking me if spending time with my friends is worth it.
♥ 1
Replying to @polx
I don't want it to be something where they decide if they follow the rules or not, we know how that ends.
Replying to @johnfsymons and @asamonek
Twitter engagement is down across all communities. You just never see the people you care about any more, and they don't see your stuff. I'm getting waaaaaay more interactions (and good ones too) on Mastodon for 25% of the followers.
♥ 1
Replying to @AnupamChander and @elonmusk
I'm pretty sure it's more or less random, though. More than once I've posted a tweet that immediately appeared with dozens of views. I assume they just dumped whatever fanciful adtech reach measure they had into the tweet.
♥ 2
Replying to @null and @shadowymaus, @mer__edith
Amazon Prime Video, with access to years of my interests across different media, insists that I MUST be wanting to watch me some NFL.
♥ 1
Replying to @null and @MadronaCitizen
It's just fucking amazing.
Hey, I know a lot of you tried Mastodon with the early fits of the petulant moron and were put off by that experience.
But as this site keeps getting worse and more dead (it's like they hired the engagement people from Meta), I'd like to encourage you to try again.
♥ 17↺ 12
I'm not going to claim that it's perfect over there, but I can share my experience.
First, picking a good client makes a world of difference, and there are now some really good ones.
Second, my timeline is now waaaaaay better over there, none of that "For You" crap.
♥ 3
I'm also seeing way more discussions and more engagement with eg. my blog over there, and this on a quarter of my followers.
I'm just missing some of you folks over there. So come join the party!
♥ 3
Replying to @mberk
I'm enjoying Tusky right now, but people are raving about Pinafore and Elk.
Replying to @faineg and @TheSharlinator
France does have some nice rice dishes, but it is also home to the notion that rice must be drenched in butter because if so much as two rice grains stick together that means it wasn't cooked well. That might be worse than not eating rice?
♥ 1
This matters because every cent that Google uses its position to divert is a cent that doesn't go to improving the Web, and this Web-wide loss of income has accumulated over the years.
In the same way that Europe underdeveloped Africa, Google underdevelops the Web. Same logic.
Quoting a tweet by @jason_kint ↗
♥ 26↺ 15
Replying to @mberk
Yes! The digital forensics involved in proving that are going to be challenging.
♥ 1
Whenever I hear someone from Google earnestly saying "we're doing this for the publishers" I can't help thinking, with sincere sympathy, about how they will feel when they realise who they're really working for.
Quoting a tweet by @jason_kint ↗
♥ 24↺ 8
Replying to @montezumachavez and @alextcone, @airgups23, @Myles_Younger, @Chronotope
That's just how I do my public speaking.
♥ 4
Replying to @montezumachavez and @alextcone, @airgups23, @Myles_Younger, @Chronotope
ParisWeb closing keynote, yes :)
♥ 3
Replying to @Chronotope and @montezumachavez, @alextcone, @airgups23, @Myles_Younger
That's nothing, you should have seen the Karl Dubost lightning talk in which he cut off the legs of his pants live to make a laptop bag (in under 5min). People were fainting.
♥ 4
See flickr.com/photos/franck-… (more of the vibe in my faves).
♥ 2
Replying to @montezumachavez and @Chronotope, @alextcone, @airgups23, @Myles_Younger, @privacypen
I haven't run that session in quite a few years, but I could be tempted to bring it back with the right set-up 😂
♥ 3
Replying to @RobertJBateman
I mean, @nickclegg isn't lying, it is consistent with their values, notably cowardice and hating democracy.
♥ 1
Replying to @Chronotope and @montezumachavez, @alextcone, @airgups23, @Myles_Younger, @privacypen
Closing keynote, being a lightning talks session!
♥ 3
Wait, did it always say "Verified" Views? Had I tried to come up with something that sounded more like ad fraud, I don't think I would have succeeded.
♥ 4
Replying to @Chronotope
It's like a whole new pit of despair. I can't imagine buyers are reassured much!
♥ 2
Replying to @null and @MadronaCitizen
You say that because you only got 3 verified views.
Replying to @null and @evyanastasia, @neighbours_wifi
Underrated tweet.
Replying to @Chronotope and @montezumachavez, @Law_Privacy
Anything that uses HTTP is pretty straightforward, not much room for fragmentation. The UI is the harder bit.
♥ 4
Replying to @Law_Privacy and @Chronotope, @montezumachavez
GPC interop is many orders of magnitude easier though.
♥ 3
Replying to @ak_1490 and @hemanthkumarak, @RobertJBateman, @siobhansolberg
Does it only include the time period for rejection? The UI looks like it could be for any decision. I think the guidance is that consent is for max one year, but there is no floor on how often you can ask, which is dumb af. One month is way nicer than most CMPs!
♥ 2
Replying to @ak_1490 and @hemanthkumarak, @RobertJBateman, @siobhansolberg
I don't think that it's possible to assert that without knowing more. That description could easily be covering DoS prevention or load-balancing, for instance.
♥ 2
Replying to @wheresaddie
I used that same background to give a few privacy talks at NYT and if that's not the right vibe for digital teaching I don't know what is.
♥ 2
Replying to @RobertJBateman and @hemanthkumarak, @siobhansolberg
So load-balancing counts; DDoS prevention is more complicated as you point out, however if I were someone's privacy person I'd point out that security measures that are defeated by the refusal of consent are useless and be prepared to make the case.
♥ 4
Basically, if the third party is a processor, has good protection measures, limited collection, short retention, etc. and if the threat is real then I'd be comfortable with it — even though those cookies are likely to be fingerprinting. But yeah, it'd prolly scare the lawyers :)
♥ 2↺ 1