Replying to @TheAnnaGat
If I stop tweeting it will be because my pile of books to read has collapsed and I'm suffocating under it 😂
August 2018
176 tweets
If you this was deemed insufficient consent, you can kiss goodbye to almost all #GDPR consent you've seen on the Internets!
Quoting a tweet by @gabrielazanfir ↗
♥ 1↺ 4
Replying to @TheAnnaGat
This actually almost happened to me already, shelves of books two rows deep collapsed into my studio. I jumped onto the bed, but the room was shockingly full of books!
Replying to @TheAnnaGat
When it comes to acquiring or reading books I have long given up any delusion of free will.
♥ 1
Replying to @TheAnnaGat
I do the converse: I lay a course for where I want my mind to go by reordering my to-read piles. The permutations are just as Borgesian, though.
♥ 2
This is a great opportunity to learn how to fight the good fight.
Quoting a tweet by @NOYBeu ↗
♥ 3↺ 1
Replying to @allymacdonald
You're the heroine the Internet needs.
It's interesting to read livetweeting from #DWebSummit and #balisage in parallel. You folks should meet. It would also be neat if you'd both stop organising when I can't attend.
♥ 8↺ 3
Browser extension that simplifies declining #GDPR consent for Quantcast’s banners (I hope this will get extended to others).
It’s almost like there should be some kind of API for this…
Quoting a tweet by @vtoubiana ↗
Replying to @carljackmiller and @TheAnnaGat
I don't disagree with the general statement, but I think the specific example is just good old Sartrean bad faith.
♥ 2
The catastrophic scenario is that the Democrats win the midterms just in time for Trump's asinine policies to start tanking the economy. By the time they can fix half of it there's a Republican landslide in 2020, despite fairer voting rules.
♥ 3
Replying to @erikphoel
I'm not entirely sure. Facebook did introduce negative emoting but I don't think it goes against the spread of negativity-based engagement.
Who's a good doggie? Huh? Who's going to kill all humans? It's a good doggie, yes!
Replying to @TheAnnaGat
I would say: ability to switch quickly between levels of abstraction, and between in/deduction as needed.
♥ 2↺ 1
Is there some good literature on retrocausality (in physics and/or philosophy)?
↺ 1
Replying to @TheAnnaGat
Now I'm having terrible ideas about Feyerabend/Phil Collins crossover songs.
Replying to @erikphoel
And tally ho!
Replying to @erikphoel and @TheAnnaGat
That does sound like an interesting avenue but I'm not sure that teleology sufficiently covers retrocausality. I'm more interested in looking closely at assumptions that causality needs to be cisluminal. I think I'll start with the SEP ☺️
plato.stanford.edu/entries/causat…
♥ 1
Right now I'm considering acquiring some land in Nova Scotia to start planning for a Plan B residence. Any alternative suggestions?
nyti.ms/2O7HS4I
♥ 1↺ 1
Replying to @null and @siilime
Good thinking. Everyone has a boat in Finland so you need it no matter what. One of my main questions is whether Antarctica wouldn't be better as a large temperate landmass. But you can't just buy land there right now, so that's a limiting factor.
Replying to @TheAnnaGat
Yeah, a few US papers approached the GDPR with fear rather than reason. Sadly the @latimes is one of them.
♥ 1
Replying to @cmhphilosophy and @TheAnnaGat
And scientists, journalists, hackers, politicians... Feedback tends to be structurally negative, it's always worth countering that.
♥ 2
Replying to @TheAnnaGat and @latimes
No, they did that on May 25th. You can VPN...
Nostalgie de l'UTF-8.
♥ 6
Replying to @null and @johncarlosbaez, @juliagalef
That's what you call the yolk of oppression.
Replying to @TheAnnaGat
I'm actually really looking forward to coming up with new recipes. And think of the wine pairings! Wine 🍷♥ 🐛
Replying to @TheAnnaGat
Oh, it is not uncommon for older French cheese to contain bugs already 🧀😋
♥ 1
Replying to @__apf__
I once tried to write a complete CSS parser using a single Perl regex (using some of the less orthodox bells and whistles). It baffles me that people don't just love these. It's like real life magic!
♥ 4
Replying to @gsnedders and @__apf__
I've thought about that but the AAA and friends make it tricky. Maybe a streaming one though?
Replying to @gsnedders and @__apf__
This could prove fun 😏 I forget how to make a pattern self-nest, my Perl is rusty, but it might just work.
Quand il fait trop chaud pour récrire un mot sur ta carte...
♥ 1
Seems legit. t.co/AHwBkRzx61
♥ 1
When dead wasps are being eaten by ants, other wasps try to intervene. I can't tell if it's to help or to eat their own. Either way they get their chitinous arses handed to them by tiny ants with formic acid.
♥ 3↺ 1
Replying to @leonor_de_b and @DisinfoEU
Ce n'est pas parce que c'est public que ce ne sont pas des données personnelles, en l'occurrence de catégorie spéciale. C'est comme ficher des gens en manif.
Replying to @ioster and @leonor_de_b, @DisinfoEU
Faux. Le lien avec des noms n'a rien des nécessaire. Ces informations se rapportent à des personnes, ce qui sont des informations personnelles.
Replying to @TheAnnaGat
I don't use that word precisely for that reason: I can never remember if I am working with the right definition or not.
♥ 1
Replying to @patcable and @__apf__
I'll take that over "you know about computers, don't you?" any day of the week.
♥ 9
Corsica is kinda hard.
♥ 2
Mobile browser vendors: developers JavaScript too much, performance sucks!
Also mobile browser vendors: I see you switched between apps, let me reload that page for you!
♥ 1
Replying to @yoavweiss
True, but gratuitously reloading a page isn't one.
An important read about Google’s continuing war against journalism and shared truth.
Quoting a tweet by @math_rachel ↗
♥ 6↺ 3
Replying to @jonquilynhill and @A_M_Dolan
Mayonnaise is immortal.
If you had unfettered legislative powers, what would regulation of recommendation systems look like?
↺ 3
Bonsoir.
♥ 4
The @wirecutter is truly awesome and they're looking for an editor in chief to make it awesomer. Someone to give it a more global and cosmopolitan bent would really rock.
Quoting a tweet by @samdolnick ↗
♥ 3
Replying to @fantasai
Hahaha, nice 😁
Replying to @jw_lockhart and @msalganik
I thought of explainer docs, but they wouldn’t help much with YouTube conspiracy content. Auditing could work, but it would need a strong framework to rely on.
♥ 1
Replying to @philfeld
Would that work against the exploitation of gullibility though? “You are seeing this list of videos because thousands of other hapless conspiracy theorists watched it too” might not work? :)
I’d love to read more!
You can judge the quality of a vacation by how long you need to remember your work laptop’s password.
♥ 11↺ 2
Replying to @philfeld
You might be right; I certainly like the idea that contextualising a recommendation to the subcommunity that originated it might be enough to change the dynamics. I hope someone tests this.
Thanks for the links, the abstracts look very much like my cup of tea.
Hey @cpomagazine, is something the matter with you that you’d retweet such vomitous grot from these losers?
↺ 1
Replying to @jw_lockhart and @msalganik
I completely agree that strong transparency requirements can really help, especially in cases in which one person assessing for many is preferable to each assessing for themselves. (Privacy is also like that.)
Replying to @jw_lockhart and @msalganik
I wonder to what degree one could require publication of metrics that would be meaningful across many recommendation systems. Eg. tendency to break down shared reality.
An excellent read from @nickconfessore for @NYTmag. If you wish to peek at the intense politics behind privacy issues, it's a great place to start.
nytimes.com/2018/08/14/mag…
♥ 5↺ 2
This is a clear and short description of the kind of influence tech companies have in political circles. They are kingmakers.
♥ 1
And this is an excellent explanation of why user consent is toxic for privacy, and unsurprisingly the preferred method of the ad industry.
♥ 4
A small issue with the article: this is not true about the GDPR. Asking for permission is just one way to do it. This small error is a shame because it's aligned with Big Tech disinformation on the topic.
But with that nitpick in mind, read the article, it's great.
♥ 2
Replying to @ubiquitic
Consent is only one of six legal bases under which you can process personal data. Other options are contractual, legitimate interest, legal obligation, vital interests, or public interest.
♥ 1
Replying to @tobie
Well, no one retweeting it doesn't help. Also, I don't have a Soundcloud account.
♥ 4
Replying to @jw_lockhart and @msalganik
Protecting recommendations from being gamed is definitely useful, but the context here is thinking about how and whether to regulate them. The threat model is more around destroying democracy, radicalisation, attentional addiction, etc.
Replying to @jw_lockhart and @msalganik, @dfreelon
Can you meaningfully ask what is being optimised for without the means to at least test it? Would a requirement to make something like algotransparency.org easier to implement be useful?
Any experience using netlify? It looks like a pretty neat solution that does one thing well, but I’d love to hear experiences before I dive in.
♥ 1
Replying to @null and @nhoizey, @jamstatic_fr
I don't feel like a new Slack, but that sounds like endorsement enough, thanks!
Replying to @FemSportLab and @SDMumford, @LogicalAnalysis
Yes, it's the dynamical correlate of a recipriversexcluson.
Replying to @TheAnnaGat and @erikphoel
I’m not sure that’s obvious. A pre-Carboniferous civilisation would not have used fossil fuels, so no plastics, etc. What would industry look like if made by plant-based intelligence? It’s a big bag of questions :)
♥ 2↺ 1
Le saviez-vous: Lio a enregistré une version anglaise de “Les Brunes Comptent Pas Pour Des Prunes” qui s’appelle “Brunettes Are No Puppets”. C’est, euh, c’est… je vous laisse juger. open.spotify.com/track/0gvxs4Rh…
♥ 1
Frankly, it’s still a hell of a lot better than what most web sites do.
Quoting a tweet by @PrivacyMatters ↗
♥ 3
Replying to @PrivacyMatters
Are you trying to get into a secret base of some sort?
Replying to @triblondon and @PrivacyMatters
I agree! There are details I don't like and I'm not sure about the data being collected in the first place, but it's otherwise surprisingly good NG
Replying to @CT_Bergstrom
Do you know if there's a transcript or video? That sounds right up my alley!
♥ 1
Replying to @CT_Bergstrom
Thanks!
Replying to @TheAnnaGat and @sknthla, @taralouisereddy, @ruchowdh, @rivatez, @JohnSigmon, @meghafon, @FeryalMP, @SimonDeDeo, @clairlemon, @MarkLutter, @C_Harwick, @IxyLabs, @ChrystalDing, @PeterIsztin
Oh, thank you, that's frightfully kind!
It's lovely to meet you all! 🤓 🤗
♥ 3
The share of plants in all Earth life is impressive.
If I think it used to be double that (and far, far more interconnected at several scales), then think about plant cognition, I can't help wonder: if they were sentient and civilised, would we notice?
Quoting a tweet by @kottke ↗
♥ 1
En même temps, le criminel qui utilise Messenger il mérite un peu de se faire prendre.
"Kikoulol! Tu veux on va braquer une banque? 🙌💕"
Quoting a tweet by @oliviertesquet ↗
Replying to @Noahpinion
Leibniz, Popper, Orwell. I also like Manu's choice.
Replying to @TheAnnaGat
I equally prefer rubidium to bromine. I doubt there are any good practical jokes involving bromine...
♥ 1
Replying to @TheAnnaGat
It is indubitably cute, but it does make me wonder what kind of library he had been assembling that didn't have a dictionary!
Where should I buy shoes in London?
Replying to @TheAnnaGat
Ah, the binding energy, indeed that is less readily available!
I would make a terrible teacher's pet, but all the same I feel compelled to read this book, since I haven't. 🤓
♥ 1
Replying to @TheAnnaGat
Touché! But there is a time factor, the current ones might die on my feet.
♥ 1
@zeynep's recent article for MIT Tech Review is (characteristically) cogent and full of gems. You've likely seen it shared already, read it if you haven't.
technologyreview.com/s/611806/how-s…
♥ 3
"Old gatekeepers failed in many ways, and no doubt that failure helped fuel mistrust and doubt; but the new gatekeepers succeed by fueling mistrust and doubt, as long as the clicks keep coming."
"The problem is that when we encounter opposing views in the age and context of social media, it’s not like reading them in a newspaper while sitting alone. It’s like hearing them from the opposing team while sitting with our fellow fans in a football stadium."
♥ 2
"Our cognitive universe isn’t an echo chamber, but our social one is. This is why the various projects for fact-checking claims in the news, while valuable, don’t convince people. Belonging is stronger than facts."
"It was to make the world “more open and connected.” It turns out that this isn’t necessarily an unalloyed good. Open to what, and connected how? The need to ask those questions is perhaps the biggest lesson of all."
♥ 2
Having spent much time and energy trying to make things better with the Internet, the sobering truth of what has actually happened over the past few years hits close to home.
♥ 2
Replying to @geoffdelc and @ICOnews, @LINCnil
Do you work with the private sector when working on privacy-related innovation?
Replying to @olivierthereaux
Thank you! Happy to grab coffee if you have a spare moment!
Replying to @geoffdelc and @ICOnews, @LINCnil
If you plan to work on something related to media, let me know!
♥ 1
I am glad to see @DCNorg's “Google Data Collection” report. It provides a nice overview of Google's data collection practices which are worse than you think, even when you think they're bad. #GoogleDataCollection digitalcontentnext.org/blog/2018/08/2…
♥ 19↺ 19
An Android phone with Chrome running in the background that is not being interacted with will send data home to Google on average 14 times every hour — or once every 4 minutes and change. #GoogleDataCollection
Two thirds of the data that Google collects about people is done passively, without the user themselves volunteering the information in any way. #GoogleDataCollection
♥ 1↺ 4
On Android, if you turn off WiFi your location relative to WiFi signals is *still* being tracked and shared with Google. You need to switch Wifi scanning off separately. See below the event reporting to Google that the user is walking on foot. #GoogleDataCollection
♥ 2↺ 3
This of course is all before you enter the world of online advertising, which spans most of the Web and apps (by usage). In typical use, Google gets information about you 90 times per hour — or once every 40 seconds. #GoogleDataCollection
♥ 2↺ 1
None of this is inescapable. It works mostly because both Android and Chrome are deliberately designed not to offer any protection for user privacy. They act as data controllers independent of the user's agency; their sharing is subject to the #GDPR. #GoogleDataCollection (fin)
♥ 3↺ 2
I’m considering opening a position on my team for someone who builds tools to make other people’s lives easier & shinier, using whatever’s handy in a heterogeneous environment. Is “Toolsmith” a good name for that? Or would you have a better description in mind?
♥ 8↺ 1
Replying to @remitbri
Not too old for me! But yeah, probably too old for the likely hires… Also, I’m not sure I could work next to that haircut.
♥ 1
Replying to @tobie
Yeah, you jest but that might be the default... The thing is, I'm pretty sure you get better data governance engineers if you hire people who don't know what data governance is 😊
♥ 3
Replying to @drogersuk
I don't know, that feels a little fake but maybe that's me?
Replying to @anthony_ricaud
Yeah, that's why I ask. But "productivity engineer" sounds like "growth hacker" to me...
Replying to @ChristosEllina1
I think I might have trouble getting that past HR 😏
I have just discovered "The Ladybird Book of Mindfulness" and it is the best thing.
♥ 3
Replying to @null and @amyvdh
I think it might be 😁
Replying to @stimpee
That’s what’s hard when trying to make an enterprisey position sound as interesting as it really is!
Replying to @podehaye and @profcarroll
Article 15 states that “The right to obtain a copy … shall not adversely affect the rights and freedoms of others.“ If revealing fraud scoring would potentially subject others to fraud, you might be able to keep it secret under this provision. (IANAL, etc.)
♥ 1
Replying to @profcarroll and @podehaye
A fraudster with sufficient resources could use black box testing to figure out what behaviours lead the score to increase by how much, and thereby try to fly under the radar.
I agree that it’s a somewhat contrived threat.
The form is a bit of a hassle but if you are a EU citizen, *please* sign & circulate the petition to grant Brits EU citizenship post-#Brexit.
Our British friends should not be penalised for not having a single competent politician. It happens to everyone.
eci.ec.europa.eu/002/public/#/i…
♥ 5↺ 6
Le formulaire est un peu pénible mais s’il vous plaît signez & circulez la pétition pour donner aux Britanniques la citoyenneté européenne post-#Brexit.
Ils ne devraient pas être pénalisés pour n’avoir aucun politique compétent. Ça arrive à tout le monde.
eci.ec.europa.eu/002/public/#/i…
↺ 2
Replying to @macsym
restorative justice > retributive justice
Replying to @alexmilowski
In most countries 🤗
Replying to @manuonzeroad and @vlentz
I don't participate to the budget and yet enjoy the rights. Besides, punishing the individual for the actions of the group is a tribalist approach I don't subscribe to.
Replying to @jaspio
Huh?
Replying to @torgo
Yes, exactly the same!
Replying to @TheAnnaGat
I'm not sure what it would mean, but "The Death of Tragedy" would be a kick-arse name for a book.
♥ 1
Replying to @WKCosmo
Or, you know, a baby human.
The idea that behaviour is not knowledge is weird.
♥ 1
Replying to @darrel_miller
Yes. People treat email like a chat system, hence the top posting and all. Two questions, two lines!
♥ 1
Replying to @jaspio
I'm not British, I have no idea what you're talking about.
Replying to @csarven and @w3c_wai, @stevenpemberton, @TzviyaSiegman
If you use RDFa, you’ll need to be sure your subject and all are properly set. It also brings in baggage like prefix processing, etc.
Relying on `role` to set a landmark is more straightforward. If you’re looking for a rabbit hole, there’s an RDF mapping for that.
♥ 1
Eight (yes, 8!) tech manager positions open at one of the coolest orgs out there, working with incredibly smart and kind people. Apply while it's hot!
Quoting a tweet by @nytopen ↗
♥ 2↺ 4
Replying to @null and @amyvdh
A few I don't recognise, most are true. I can't believe anyone knows about Alizée outside France. "Ella Elle l'a" is actually a cover of French singer France Gall.
This. This is exactly how I die. The scary thing is that I narrowly avoided it once already.
Quoting a tweet by @PhilosophyMttrs ↗
♥ 10
Replying to @TheAnnaGat
I tried the Kindle for over two years but returned to paper. It just doesn't stick to my brain. 🧠
I had a nightmare last night that I was forced to move into one of those impractical Hausmannian apartments and couldn't keep my books; I sympathise...
Replying to @roessler
Ha! A great mind no doubt!
Replying to @gsnedders
A bit over ten years ago I was awoken by a weird rhythmic creaking in my tiny studio. It was my bookshelf that had weakened in a spot and was getting unstable. It suddenly broke and filled the room with books, I jumped onto the bed to safety!
♥ 1
Hey twitterers: any recommendations for a tool to visually edit dependency trees (repurposing something made for other uses is fine)? Failing that, any good DAG visualisation you like? Thanks!
In the US those days multiline wall calendars to organise one's family (which I know as "Family Planners") appear to be called "Mom Planners". #patriarchy #wtf
♥ 1
Replying to @recifs
I was afraid someone would say that 😏 I have some difficult memories of trying to make spaghetti diagrams look good...
Replying to @dimitribosch and @progapandist
No need to learn about politics, you can just interact directly with the CSS WG. The thing you need is to do your homework though. People have looked at form styling a lot, there are going to be reasons why it's not there yet.
♥ 2
Replying to @TheAnnaGat
And you outperform those who prepared for the wrong thing. Beware survivor bias!
♥ 1
"The Wind Cave", by Haruki Murakami newyorker.com/magazine/2018/…
Replying to @null and @_james_fuller
Gephi is almost certainly the wrong tool, but the other two look interesting — thanks!
Replying to @barefootliam
Trees are easy, I could even just get folks to use an outliner for that. But the real data is DAGs and without that it’s going to be a lot of copy/pasting.
The @nytimes, rocking the blockchain way, way before it was cool.
motherboard.vice.com/en_us/article/…
♥ 8↺ 5
Firefox finally joins Safari and Brave in the list of browsers that care about their users more than about advertisers.
Hey but there's some browsers missing from that list! Huh. Can anyone think of a reason why? 🤔
blog.mozilla.org/futurereleases…
♥ 29↺ 13
*Great* article from @rachelnabors about the need for browser diversity. I do however wonder if she deliberately picked mitochondria to represent JavaScript based on their key role in apoptosis 😏
Quoting a tweet by @css ↗
♥ 9
Replying to @trekonomics
There's a bunch of relatively recent evidence that multicellularity might actually be much easier than we thought. The transition has been made several times in the lab (granted on modern unicellular life, but still).
♥ 1
Replying to @trekonomics
I compressed two ideas in my response a bit too much. I don’t know about reproduction of the eukaryotic transition; I recall reading about evidence that it happened more than once but I can’t trace it down.
But I don’t think you need eukaryotes for complexity: multi is enough.
♥ 1
Replying to @trekonomics
And here is an interesting read I could trace down from a couple of months back on reproducing multicellularity: sciencemag.org/news/2018/06/m…
♥ 1
Replying to @trekonomics
I’m annoyed that I can’t trace it back; it’s not this, but at least the cyanobacterium example in here was mentioned (but it’s cool, I didn’t know there were yet other endosymbionts like that) en.wikipedia.org/wiki/Endosymbi…
♥ 1
Replying to @BrendanEich and @xpost1, @prywatnik, @brave
I know Google does deals in which they strictly forbid switching privacy settings on by default, but am not sure of such a deal with Mozilla.
Do you have documented evidence for Bing? This isn't without consequence.
Replying to @BrendanEich and @xpost1, @prywatnik, @brave
I wish there were documented evidence. Using your clout in search to force the behaviour of competing browsers and platforms is not a minor antitrust violation...
Replying to @BrendanEich and @xpost1, @prywatnik, @brave
I am well aware. But reporting illegal activity is not a breach of confidentiality, as far as I know. Not to mention mechanisms for whistle-blowing.
Replying to @BrendanEich and @xpost1, @prywatnik, @brave
I wouldn't assume you are. Some of our mutual friends, however...
Replying to @BrendanEich and @xpost1, @prywatnik, @brave
The answers seem to largely think that "mutual friend" is better if not the superior option!
Is anyone else getting this message from the @washingtonpost when they’re not blocking ads? I’m blocking tracking, but definitely not ads.
♥ 2
Wait… The @washingtonpost is calling Firefox’s tracking protection an “ad blocker”.
There’s room for disagreement on a lot of things, friends, but it’s hard to see this as anything short of a shameless lie.
washingtonpost.com/steps-for-disa…
♥ 3↺ 1
Replying to @hillbrad and @washingtonpost
Except that is not a faithful representation of my position. “We need to track you to make money” and “we need to advertise to you to make money” are two very different propositions, and conflating them is part of the problem.
♥ 3
Replying to @hillbrad and @washingtonpost
I know that, in fact it’s my job to know that. My employer is intimately familiar with the issue.
This does not change the fact that conflating advertising and tracking is disingenuous. It feeds the Google/FB/adtech narrative that this is the only/right/best way to do this.
♥ 1
Replying to @hillbrad and @washingtonpost
I would *happily* give a lot of my personal data to the WaPo. I trust them. That's why I'm disappointed that they would so openly misrepresent the facts. Us publishers have to live with adtech's practices to a point, but we don't have to spread their lies.
Replying to @hillbrad and @washingtonpost
I can hear that argument for the CTA — this however is a whole page of it: washingtonpost.com/steps-for-disa…
If Jean-Claude Juncker has any smarts this is the time to offer Canada EU membership.
I’m only half joking. Think of the mutual wins.
Quoting a tweet by @brianleroux ↗
♥ 4
Replying to @benadida
Note that the stats I'm seeing with respect to DNT being enabled indicate much higher user interest than ever acknowledged in the Web community.
I think it's time to look at DNT again.
♥ 4↺ 1
A good thread about how adtech's refusal to consider sensible, negotiated options is leading to an arms race they will lose.
Quoting a tweet by @benadida ↗
♥ 17↺ 7
Replying to @AmeliasBrain and @washingtonpost
In an open exchange, you get much, much higher CPV ads if the user is recognised than if they are not. When an ad is shown literally hundreds of companies (not serving that ad) get to know that you were visiting that page.
♥ 1
Replying to @benadida
This is preliminary, we're digging more, but in the 20% range. One surprising number is ~12% of Chrome users despite Chrome beings pretty good at hiding the option. This has to include non-technical users. It shows people care.
♥ 10
Replying to @stimpee
On the contrary, think about the exports, notably of sanity! And the hockey games!
Is there a running app that's not about eating your privacy? Happy to pay!
♥ 3