It looks like the NYC DOE is cancelling Spring Break? I can't think of a time when teachers have needed a break more, it makes no sense.
♥ 13↺ 1
April 2020
232 tweets
Replying to @laurendw
The meal support system can operate without teachers, in fact it's been doing so for a few weeks now!
♥ 1
Replying to @EvanSelinger
We're already all Zoom and Google Classroom at ours. I haven't yet managed to tell my fifth grader that she might not see her group of friends again before moving on to middle school 😑
♥ 1
Replying to @alex_kuhner
I think the kids and us will be OK (though a break would not hurt) but the teachers have been having a hard time, being frontline tech support for parents and picking up really bad software they have to use every day.
April Fools' prank for me this morning was a fake book with a great title, as well as that cute little octopus on my back! 🐙 📙
♥ 9
If you want to run your own IT stack instead of giving away all your usage data to a potential competitor, but still want someone else to run it for you, this looks like a pretty damn good option: startupstack.tech.
♥ 4↺ 3
Replying to @gsnedders and @ericlaw
LOL, if by "really sloppy" you mean "no one ever thinks about them ever again" then yeah 🤭
♥ 1
Replying to @mariafarrell
Event just reading is proving challenging…
I initially read "quantum exorcism" and my brain has a fun little moment grappling with the implications before finding the correction.
Quoting a tweet by @torschau ↗
♥ 3
Replying to @kottke
“Some people aren’t on this list.” That statement is false the moment you include it in the list :-)
Set-theoretic jokes apart it’s a great list, thank you.
Replying to @_alastair and @justinph
We recently joined the Privacy CG (haven’t started participating because that happened at the same time as The Current Events, but will soon).
I have a draft document on why/whether we should join — very happy to chat more about this.
♥ 1
Leadership is particularly important in times of crisis. "I'm choosing not to do it," said President Trump. "It's only a recommendation."
♥ 4↺ 5
We have decided that our entire cultural production should be controlled exclusively by what engineers know how to measure.
As an engineer, let me tell you: we don't know how to measure jack shit but poor proxies for the real thing.
Quoting a tweet by @swodinsky ↗
♥ 57↺ 14
The people who decide if your work will be found (search, social) and how much money you make when someone finds it (ads) are the same small group.
The resulting groupthink means that only what can work in their view of the world gets to live. Cultural impoverishment ensues.
♥ 15↺ 2
This is made worse by a deeply-seated belief that what they've set up is scientific and rational.
Sadly, much of that world has a cartoonish, reductive understanding of science, of the "learn digits of pi" variety.
♥ 13↺ 1
Running A/B tests doesn't make anything rational, or even grounded in data.
No A/B testing will tell you you're asking the wrong question, it will just happily deliver the answer prompted by your unexamined assumptions.
♥ 31↺ 4
This is not an anti-engineering view. We need to reclaim computing from the people who thought Java, real-time bidding, or AMP were good ideas.
Maybe that is what we should call computer lib!
♥ 20↺ 5
Replying to @v0max
100% agreed! Engineering is a craft, treating its methods as epistemic leads to results that would be hilarious if we didn't put them in charge of anything important.
♥ 4
Replying to @v0max
Yes, and of high modernist ideology. It's not just ignorance, it's also thinking they're making the world a better place!
♥ 1
Replying to @null and @theantitheory
Oh. How does it get selected then, magic?
I remember the Internet before it was captured by platforms. It was better.
How tech organises work is also relevant here. People look (rightly) at gig worker but the intent is to get developers to be interchangeable too.
Tech today is entirely focused on scale, and you can't scale creative, local knowledge.
Quoting a tweet by @dmarti ↗
♥ 4↺ 1
Scale sounds cool until you realise that it can only function by obliterating differences at ground level. That's how you end up with Pulitzer-winning outlets and the 700th phone light app all being "publishers", put in direct competition as interchangeable ad inventory.
♥ 3↺ 1
A healthy digital future goes through checks and balances to scale at all levels: privacy, CDA230, unions, fiduciary duties for user agents, transparency of ad intermediaries.
♥ 4↺ 1
♥ 2↺ 2
If we're all going to wear masks for the foreseeable future, we should start using Belter hand signs as good replacements for facial expressions.
♥ 8↺ 2
Replying to @steltenpower
From The Expanse (in the novels at least).
Replying to @natashanyt and @janethaven
The worst part is that Google Classroom is a terrible piece of software. No one can make that old claim here that privacy is being traded for convenience.
♥ 3
Replying to @susanthesquark and @HuffPost
It's sad to say but they probably don't know it's there. Outlets that rely that heavily on programmatic are basically open targets, and any protection they add is likely to decrease their revenue. As advertising collapses, we have to expect this to get worse.
♥ 1
Replying to @janethaven and @natashanyt
I have it opened on two different browsers on the device my daughter uses plus once more on my laptop just so that we can work around their bugs. It's maddening.
I wonder if anyone has quantified the value of schooling just from liberated labour.
This is a very interesting thread about why epidemic models split into best case and worst case without much room in the middle.
Quoting a tweet by @CT_Bergstrom ↗
♥ 4
Replying to @PhilScottMRTPI and @MarkDriver1968, @northumbrian_, @JamesMelville
New Zealand is a highly liberal society. Competent leadership makes all the difference.
Are any of you on a solution that works for classroom video calls? Zoom has been forbidden by the NYC DOE due to legal concerns over privacy, Google Hangout/Meet is a trainwreck with people dropping constantly, MS Teams doesn’t even seem to start for some… It’s a mess!
♥ 4
Replying to @lhirlimann
Are you using it for classes? I know plenty of options exist but I’m looking for experience reports because video with 30 kids presents its own challenges!
Replying to @wayneblodwell
But have you tried them in the (particularly difficult) setting of running a first grade class meeting?
Replying to @scottbw
Have you tried it?
Replying to @rigow
None of this seems to be a good option for a first grade classroom!
Replying to @kenarchersf
Yeah, but the NYC DOE has forbidden it!
Replying to @odinho
Thanks! I’ll suggest that to the teacher :) Ha! You’re on the site :)
Replying to @odinho
I don’t think it’s local, many families left NYC and are all over. Also, it worked fine with Zoom. I think part of the problem with Meet is Google’s been having infra issues, and part because like most G. products it’s just designed to be used by G. engineers rather than users.
Replying to @odinho
Trimming one’s beard is overrated ;-)
Replying to @podehaye
Isn't iOS a clear indication of wealth anywhere? I would expect Android to overindex on covid impact.
Replying to @null and @vpdjgr, @podehaye
It would counter it some, but I'm not convinced that it would counter it completely. AFAIK, having an iPhone is one of the best predictors of being in the $100k+ category. People get richer with age, but not to that degree.
Replying to @mbeisen
This one seems pretty cool, it's basically a long weekend that's all party. I'm calling it a win too.
Quarantip for parents: if you run out of grown ups to hang out with, you can build yourself new ones!
♥ 28
Replying to @jsiegl and @PlayWell_LLC
Thanks, will look. Self hosting might be tough for some schools, though.
♥ 1
Replying to @LeonieWatson and @casinclair
Will pass along when I get to my desk, thanks!
Replying to @NoWAYNE3
I’ve used Moodle before, I don’t think that’s a period of my life I would enjoy revisiting.
Replying to @null and @sivavaid
In general, for much of learning, I agree. But looking at my kids (and granted the peer-review on this is my wife) they do need _some_ sync, maybe 30min. Seeing the teacher and the other kids in the class is energising and makes the day a lot less lonely.
♥ 1
Replying to @null and @sivavaid, @funnymonkey, @Startup_Stack
That is simply not a true statement. It’s hard, and the way the current tech economy is structured you can never do it at scale, but you can run a small ethical tech company and make a living from it.
♥ 1
Replying to @null and @sivavaid
I think classes can make informed decisions here. The NYC DOE has been doing the functioning government thing (which people in this country call “socialism”) and distributed 300,000 tablets. It seems to be working.
Replying to @null and @funnymonkey, @sivavaid, @Startup_Stack
Same! And I worked for a small but ethical startup (now a small regular company) that’s about to enter its twentieth year.
You miss out on easy bucks but you also get very loyal customers and street cred from it.
Replying to @null and @funnymonkey, @sivavaid, @Startup_Stack
I’d also like to think that The Times is a tech company these days, and that while we’re far from perfect we’re definitely striving for ethical tech solutions and continuously improving.
I don’t think defeatism is empirical here, or, in fact, ethical.
♥ 1
Replying to @null and @sivavaid
Well, I can’t seem numbers at scale yet but in my kids’ school, which has high income disparity, I’m seeing kids coming online who weren’t and classes being complete that weren’t. Of infrastructure problems this isn’t the hardest!
♥ 1
Replying to @valentinebaud
Carrément!
I look forward to the CSS module, @tabatkins & @fantasai.
Replying to @swodinsky and @larakiara, @heyitsnoah, @aripap
It is adtech. Sorry Lara. The truth is hard.
♥ 2
Replying to @AmeliasBrain and @tabatkins, @fantasai
When the project finishes we should actually do that 🤭
♥ 1
I have been thinking along similar lines as this thread (h/t @djleufer & @mariafarrell) but with somewhat different views. I think that ML has its own somewhat novel position in the "Seeing Like A State" framework. 🧵
Quoting a tweet by @Juniperbevensee ↗
♥ 10
ML doesn't *create* the legibility. It can only operate if there is already data, so the legibility has to already exist.
It does however make the data more actionable such that you can mould the world to match the utility function that trained the ML.
♥ 3
This matches the James C. Scott framework since he discusses the State changing the world back. You start wanting to organise the world's knowledge, you end up organising the world to match your knowledge model.
Building from a utility function is also very high modernist.
♥ 4
One distinguishing aspect of ML is that it makes the State more illegible to its subjects. This strengthens the power imbalance beyond what Scott discusses and I think is novel.
♥ 4
Another novel aspect is that ML enables scale with fewer natural checks and balances (from physical realisability) than anything in Seeing Like A State. And scale without checks & balances is inherently bad.
♥ 5
That said I reach similar conclusions to @emmibevensee that ML is not inherently bad but rather requires adapted governance. In fact, ML can be used *with* mētis. In fact there's an upcoming @timesopen post on a sociotechnical system associating ML with people for good results!
♥ 4
Replying to @henryfarrell and @mariafarrell
Are there drafts that would fortuitously happen to be lying around?
Replying to @henryfarrell and @mariafarrell
Also beating myself up for not thinking of such a great title.
♥ 2
Replying to @Juniperbevensee and @emmibevensee
It's an interesting point! But I don't think that they are strong or near enough to serve as strong controls. If you can run detailed invasive profiling and ad bidding for a couple of billion people like Google does, the limits on ML damage aren't strong, no?
Replying to @Juniperbevensee and @emmibevensee
Also, I mean, what do you do when you hit computational complexity issues? You use heuristics, like any good engineer! That will make the system "work" but adds yet more unexamined assumptions and illegible behaviour.
♥ 1
Replying to @Juniperbevensee and @emmibevensee, @timesopen
I think the illegibility of ML is a key difference with Scott. ML already gets a moral free pass because people don't ascribe agency to it (and don't blame its creators), and because it makes subversive action much harder to direct properly since we lack intuition for what works.
♥ 1
Replying to @Juniperbevensee and @emmibevensee, @timesopen
It's the local, practical, in-the-trenches knowledge of practitioners that Scott talks about when he looks at what farmers know compared to what the State perceives.
♥ 1
Replying to @henryfarrell and @mariafarrell
I liked that one too! But my interest is more in how big tech platforms resemble states, Taylorise other companies, etc. and the FSM parallel might work better with that than for authoritarianism and AI.
♥ 1
Replying to @null and @shauna_gm
Absolutely agreed that it's not a 1:1, however I think Scott makes a very good case that legibility is *not* understanding either! If you want to tax, legibility will bias to what can help tax (eg. data about land size as proxy), but you won't understand how people grow food.
Replying to @null and @shauna_gm, @henryfarrell, @mariafarrell
Might you share? And yes, very, in how they run taxation-like schemes too, like digital advertising. Agricultural policy is also instructive here. Publishers invest in SEO, social, and paid content (free work for platforms) in hope of traffic to monetise through platform ads.
Replying to @null and @shauna_gm
Actually, I think that with that understanding/legibility distinction I've talked myself into agreeing with you that ML does add legibility 🤭
I tend to think of it as more action-oriented but that actually matches legibility-without-understanding well.
NYC schools are not reopening this year...
♥ 1
Replying to @lhirlimann
School, so far.
Replying to @null and @shauna_gm
Agreed, data has constituencies and legibility depends on what actions you can undertake.
Replying to @Juniperbevensee and @emmibevensee
Yes, but part of my concern is that ML is hard to read for humans and that constrains our intuition for attack vectors when resistance is required.
♥ 2
Replying to @Juniperbevensee and @emmibevensee, @mutual_ayyde
I think the concern here is that it's complexity vs complexity, which is new.
♥ 1
Call me soft-hearted if you will but in my book all Fridays are pretty good.
♥ 5
Replying to @null and @shauna_gm, @kvnweb, @seanmmcdonald
Would love to hear more too! The refusal from tech to be "political", by which they mean having any kind of opinion in the face of complex issues (I think that matches your "theory-less" which is great), does seem to be contagious well outside of tech.
Replying to @seanmmcdonald and @shauna_gm, @kvnweb
Am I correct in understanding that you are looking at the agency of government here, and not of their constituents? It seems like a really powerful angle I hadn't considered.
Seine-Saint-Denis Style.
♥ 2
Bon, je sens bien que mon humour ne verra pas son génie reconnu de son vivant...
♥ 1
Replying to @tbernard1979
Voilà voilà, le jeu de mot est là.
Replying to @olivierthereaux
C'est pour les Happy Few comme disait Stendhal 🤭
(Je vais être relou.)
Vous l’avez?
♥ 5
Replying to @julien51
INDICE!
Qu’y a-t-il dans l’écran? Quel type de voiture est-ce?
Replying to @null and @nhoizey
Zoom. Dans une Benz?
Replying to @julien51
Zoom-zoom-zang!
Replying to @goetsu
Nooooon, mais tu te rapproches. Que fais-je dans la Benz?
Replying to @null and @nhoizey
“Browser Hygiene for Better Privacy - Think of it Like Washing Your Hands Online!”, by @funnymonkey
funnymonkey.com/2020/browser-h…
♥ 2
Replying to @null and @nhoizey
Wow! Qu’on aime ou pas je ne pensais pas qu’on pouvait avoir traversé les 90s sans connaitre, désolé!
Replying to @null and @nhoizey
Haha, figure-toi que moi non plus, mais je n’y ai quand même pas échappé :-D
Replying to @fgeorges
Wesh! Et toi 😁
A more perfect Union?
♥ 2↺ 1
Replying to @fgeorges
RESPECT!
Replying to @fgeorges
Tu veux dire que je t'ai mis la fièvre, en somme?
Replying to @slightlylate
It's not necessarily a bad starter bug that gets you to know the whole damn codebase 🤭
♥ 1
Phil Dowe explaining Bell inequality through an analogy involving identical twins from Sydney being flown in great numbers to Hobart or Dunedin to see how many die of "freezerphobia" might not be the clearest in metaphysics but it deserves a prize of *some* sort. ❄ 👬🇦🇺
Replying to @can
I am not a game person so my judgement may not be worth much but I saw someone playing it a decade or so ago and I have rarely seen something this boring. Cute, but boring.
The @latimes lost 1/3rd of their ad revenue. How much of that goes to adtech intermediaries? You can bet it's more than a third.
Are those costs transparent? No.
Competitive? No.
Worth losing our press over? No.
A cleanup of that market is way overdue.
Quoting a tweet by @marcatracy ↗
♥ 11↺ 2
Replying to @null and @theantitheory
That won't solve the core problem.
Replying to @v0max
Where do I sign up for that spam?
1. Lol (again) Google Classroom.
2. I once got a job that way. Aced multiple-choice coding questions in several languages, including C++ which I couldn't use if my life depended on it. 🤭
They wanted someone who got web tech better than they did so in that sense it worked out!
Quoting a tweet by @bgalbs ↗
♥ 7
Replying to @frivoal
I have a good book the title of which will surely come back to me any time now... (It's at home.)
Generally, the idea is to only lock someone up if there is no alternative whatsoever and the risk is deemed real.
For the rest, use rehabilitation, restaurative approaches, mental health care, house arrest, community service... There are options! The current system just tortures people to produce more criminals.
Replying to @frivoal
Yeah, I don't know that getting to zero prison is realistic in our lifetimes, but you get to something that doesn't look anything like what we know as prison.
If anyone ever needed convincing that DoubleVerify is snake oil…
Quoting a tweet by @nandoodles ↗
♥ 4↺ 1
Replying to @acfou and @DweebMarketing, @nandoodles
Also that is *really* not something that should be solved at that level.
♥ 3
Replying to @publictorsten
The clouds is DoubleVerify replacing the ad creative with their image because the context was deemed brand-unsafe.
♥ 1
Replying to @publictorsten
I was joking :)
I don’t have numbers to back this but I think you’re much more likely to get it if you use Chrome. You need to be tracked so you get higher CPM programmatic where someone might have bothered with brand safety.
♥ 1
In which @vicmsong tells it like it is.
If you're in the "but Google only takes 30% crowd" this is a good intro. gizmodo.com/google-does-th…
♥ 3↺ 1
This is a tale from when we moved to Brooklyn and the girls were little. I can't tell you what happened without spoiling the story but we found out later: they knew, but they didn't know what to do about it either.
berjon.com/the-confusion/
Quoting a tweet by @MarianKeyes ↗
♥ 9
Replying to @Rich_Harris
Thank you 😁
Replying to @brucel and @Rich_Harris
Pray do tell.
Replying to @brucel and @Rich_Harris
Classic. Having done the same, I can absolutely see how that happens 😁
Replying to @brucel and @Rich_Harris
I’m sure they loved it!
My first byline! I wish it had been under less tragic circumstances, but I’m still stupidly proud of my tiny contribution. It’s a real privilege to be able to provide even a small bit of assistance to such an awe-inspiring team.
nytimes.com/interactive/20…
♥ 33↺ 2
Replying to @CT_Bergstrom
I just did tiny bits, the way they have this set up is impressive. I've seen scraping operation but these people are good.
♥ 1
Replying to @hober
Thanks! :-)
Replying to @AlGalilee
C’est un système qui met les données à jour aussi régulièrement que possible, sachant que c’est un bordel sans nom parce que personne ne fait pareil et peu font bien.
♥ 1
Ah, j’ai envoyé avant de finir!
Merci, et gros bisous!
♥ 1
Replying to @null and @amyvdh
Thank you :-D
Accessibility friends! I don't know what's behind this but it's almost certainly good. Please help if you can!
Quoting a tweet by @adrjeffries ↗
♥ 1
I have never cared for video games or owned a console of any kind, but it’s becoming really hard to get my 10yo to have any physical activity (I think the lack of social aspect makes it boring). I think exercise games might help — any recommendations?
♥ 2
Replying to @tabatkins
Thanks! Will see if I can do anything about that.
Replying to @null and @benlikestocode
Thanks! Yeah, we tried Just Dance on the Mac but it's a total shit show. It refuses to work without pairing to a phone, but pairing never finds anything. Comments on the App Store indicate it's been that way for a long time...
Replying to @annbass
Thank you!
Replying to @danbri and @BeatSaber
Looks neat! But yeah, I have none of those and it doesn’t look like they can be ordered any more than a Switch can!
Replying to @donohoe
Thanks! It doesn’t look like getting an Oculus is within the realm of what’s possible, though :(
Replying to @odedshr
That looks extremely cool, but it doesn’t seem like it would help with physical activity!
Replying to @null and @mpigliucci
For a second I thought you were putting this up as one of those "reproduce this painting" challenges.
♥ 1
Replying to @odedshr
Thanks! But we have plenty of video things, they just don’t seem to be very motivating, I’m looking for more interactive.
Replying to @simevidas and @tabatkins
Thanks!
Replying to @jyasskin
Ooooh, I played this in Tokyo many many years ago! What do you recommend as the best way to run it?
Replying to @jyasskin
Yes, I’m reading up on options. People have OPINIONS! ;-)
It seems to be only available on things that are hard to buy right now, but maybe I can find a second-hand one.
Or… I could just go buy AN ARCADE VERSION ;-)
♥ 1
This is a shocking story.
nyti.ms/3biSzxr
♥ 2↺ 1
Replying to @null and @TheElleChase, @JoEllenNotte
I've often gone to unhealthy lengths to avoid having to do something over the phone, I liked the world in which people used it less 😁
My 10yo put up a rainbow in support of health workers like many other kids, but the rainbow she picked is actually a maths problem and she insisted on posting the calculations as well. 🤩🥰
♥ 23↺ 1
Replying to @clancynewyork
I don't know if that makes me a good dad but I sure have good kids! 🙂
♥ 1
Replying to @AmeliasBrain
Literally yes!
Replying to @cwarzel
I'm curious: what makes you say it's not fixable? There are at least a couple of potential changes that haven't been tried, no?
Replying to @clancynewyork
Thank you 🤗 Take care too!
Replying to @jaffathecake
I wouldn't rule out underground, the way sound travels can be weird. I once lost my cat and the stupid little cow had climbed inside a car engine from underneath. I could hear here meow but only from some angles and she was very hard to place, it took me a long while to find her.
Hang in there, I hope you're reunited soon!
♥ 1
Replying to @jaffathecake
When I finally narrowed down that she was in a car, I crawled as close as I could under it and she came out. If he's shy you being close enough might work?
Replying to @jaffathecake
This might be a ridiculous idea, but just in case: if you can get him to meow again when all is silent, put two mics well spaced out and both recording, then use delta of meow time to triangulate?
♥ 1
Greater Internet use is not associated with greater polarisation, but greater use of social networking is.
Quoting a tweet by @jayvanbavel ↗
♥ 5
Not directly related, but is there someone @mitpress who'd add @Bookshop_Org to the "Buying Options" on there?
♥ 1
Replying to @aexm and @megancgraham
One thing I've been wondering about is how that's not aiding and abetting fraud.
Great examples of Criteo, Google, etc. profiting directly from ad fraud.
These are solvable issues - in some cases tech solutions even exist - but the only thing they would get from solving them is less money, so they literally have no reason to.
Quoting a tweet by @megancgraham ↗
♥ 9↺ 5
I agree that some things are bad, but legal scholarship that typesets its footnotes as endnotes is, my friend, invariably, indubitably worse.
♥ 2
There is a proposal on the table for "First-Party Sets", or the ability to treat disparate domain properties as a single first-party, but what I would like to see is the opposite: First-Party Splits, or the ability for users to prevent reuse across services of a same domain. 🧵
♥ 4↺ 1
The use case is this: some services are offered under a single domain's umbrella, but from a user's standpoint they are distinct and data sharing between them is a violation of privacy taht should therefore be strictly controlled, ideally prescribed.
To give an example:
♥ 2
You don't want to be targeted with sweet tooth recipes on cooking.nytimes.com because you solved a candy-themed crossword exceptionally fast on crosswords.nytimes.com.
(Note: believe it or not this example is fictitious. It may, however, resemble real-world cases.)
♥ 1↺ 1
The hard part is: how do we make this work knowing that the first parties in question might not collaborate (otherwise they would simply not do it)?
Part of the answer is that you need policy to back this with teeth. The GDPR helps here with purpose limitations.
♥ 1↺ 1
But beyond policy I find it hard to support this with tech. Browsers could membrane off various parts of the same site but that feels limited.
Is this anyone has given thought to?
♥ 2
Modernised classics: “Half the money I spend on adtech is wasted; the trouble is I don't know which half.”
♥ 5
Replying to @asutherland and @annevk
Oooh, I hadn’t seen origin isolations, that’s very interesting thanks! And yes — it would need to be paired with identity separation as well as policy enforcement. Cool!
♥ 1
Replying to @dveditz
Yeah — cookies make everything painful. That's one reason I'm very tentative about this, I think it will require a fair bit of plumbing work.
♥ 1
Replying to @null and @Joyandfive
Thanks a lot!
These stories make me wonder how long it is before one of these seizures turns into a shoot out, somewhere where things are bad, between local and federal agencies.
Quoting a tweet by @NPR ↗
♥ 3
Replying to @donohoe and @jason_kint, @thori, @distractal, @SenMarkey
In fairness, Google Classroom hardly even works on Chrome either. I'm used to bad UX from Google but this one feels even more like engineers thinking they don't need designers than usual.
♥ 1
Beyond the privacy and harm to innovation, one really painful effect of monopoly is how much we end up having to put up with crap products.
At this point when I see Material Design I know I'm about to relive the Windows 95 experience.
Quoting a tweet by @profcarroll ↗
♥ 3
Replying to @rigow
I don’t know about speed, most people seem to prefer Zoom simply because it has better UI. Meet is, uhm, pretty bare bones. I still haven’t figured out how to go from the meeting landing page back to home without editing the URL.
Replying to @rigow
Don’t get me wrong, though, there’s plenty in Zoom I don’t like. But the fact that it took a pandemic for Google to notice that people don’t like Meet is very indicative. The problem with being a monopoly is your products are successful even when they suck, no feedback loop.
Replying to @annbass
Agreed. They have jurisdiction to investigate but seizing seems like it would be on shaky legal grounds (not to mention ethically dreadful). And no one knows what's done with this stuff.
♥ 1
Replying to @john_lichfield and @geoffdelc
At the national level, yes, at the local level they seem to have given up on getting confirmed cases data?
My 6yo invented a cunning cat trap: a highly realistic depiction of a mouse with rolled lightly sticky tape on it ("so you can get the cat off when you catch it").
♥ 6
It's not entirely clear as this tweet heads for publication that the beast is shaking in her stylish yet affordable boots.
♥ 4
Replying to @annbass
We do! I don’t know if that one can be kept with the stickies, but for the rest yes :)
Replying to @kyotonio
They’re not intimidating, they’re cute and fluffy, even if talking about privacy can make them grumpy ;-)
♥ 1
Replying to @null and @SimonDeDeo, @JimBaggott, @mpigliucci, @neuropoetic, @seanmcarroll, @skdh, @johncarlosbaez, @litgenstein, @JenLucPiquant, @nattyover, @DavidDeutschOxf, @DavidSpergel
I'm not sure how you'd account for the time lag between metaphysical discussion or measure improvements to physics output, but you could probably detect something like that meeting in twitter exchanges 🤭
Remote meetings are more exhausting than face to face ones. Some good notes in this thread.
Quoting a tweet by @add_hawk ↗
♥ 8↺ 6
Replying to @mariafarrell and @Annaleen
Hahaha, yes! The ending is *so* satisfying in how it opens the door to more of those. "There was actually a job that combined my skills as an academic and a murderer." I mean, where do I sign up?
♥ 2
Replying to @Annaleen and @mariafarrell
That's one way to make the preprint debate more salient.
♥ 1
Replying to @WKCosmo and @REasther
That is actually what Little Pony entrails look like after the tornado is done tossing them around.
♥ 2
Replying to @Gnppn
Already being looked at, merci!
Apple and Google have been avidly misusing people's data, and enabling the misuse of data on unprecedented scale, for a long time.
And now that data might come in useful, people don't trust them. *shock*
Quoting a tweet by @drewharwell ↗
♥ 45↺ 18
If there's one number to take away from that article, it's that people trust Apple/Google less than they trust American health insurance companies.
♥ 9↺ 1
Replying to @suyashcjoshi
There can be several! Apple came up with the IDFA and still exposes it.
Replying to @johnwalters_sd and @jason_kint
IDFA has been enabling massive misuse, and it's their invention.
People who are worried for their lives and livelihoods will take risks, the number should be way higher. Look at the comparison: they trust them less than health insurance...
♥ 1
Replying to @getwired
IDFA is the only citation needed.
Replying to @Moonalice and @suyashcjoshi
Sure! But IDFA is a pretty big hole there.
Replying to @djflore
IDFA.
Replying to @djflore
Comparing to Google isn’t very useful, at this point even the NSA is better for privacy than Google! :)
IDFA is a universal ID, that’s a couple of steps down from cookies. It’s functionality engineered for the single purpose of privacy violation. I’m not impressed!
♥ 1
Replying to @djflore
Additionally, IDFA vs ITP is a clear indication of Apple’s strategy to demonetise the Web to force people into sharecropping its app store. Not a privacy issue in itself, but the IDFA/ITP difference shows a very cynical approach to privacy.
Replying to @v0max and @Moonalice, @suyashcjoshi
If that’s the intent, why not yank access to hardware IDs and establish an identity storage mechanism matching the restrictions of ITP?
Replying to @v0max and @Moonalice, @suyashcjoshi
I know, but nothing keeps them from updating it to match ITP restrictions, and yet they choose not to. The end result is highly invasive tracking for the majority of users while differentially demonetising the web in favour of their fiefdom.
♥ 1
Is it better than Google? Of course. But IMHO it's still largely privacy theatre - transparency and choice - that just so happens to further their anticompetitive practices. I'm not surprised that would earn them distrust?
♥ 2
Replying to @v0max and @Moonalice, @suyashcjoshi
For clarity, I didn’t mean to imply they hadn’t done hardware IDs — I was drawing a parallel “IDs+IDFA” / “IDs+something better”. Sorry I was unclear.
Replying to @v0max and @Moonalice, @suyashcjoshi
Don’t get me wrong, what they’re doing on the Web is helping and I’m thankful for that. But the credit (and criticism protection) they’re getting for that feels a lot like celebrating the scorpion for riding halfway across the river with the frog.
♥ 1
Replying to @v0max and @Moonalice, @suyashcjoshi
I’m certainly sensitive to the idea that good actors can be forced into difficult balancing acts…
That said, I don’t buy that they need to compete for devs. Their marketshare and the wallets of their users are more than enough. They’re competing against he Web, for the market.
J’ai l’impression que le projet OpenCovid19-fr qui sourçait les données COVID pour la France est au point mort (il y a juste une collecte automatique à laquelle manque par exemple les cas confirmés hors national).
Il y a une source alternative?
github.com/opencovid19-fr…
↺ 2
Replying to @johnwalters_sd and @jason_kint
Yes, that would be interesting. Trusted less than the health insurance industry is a very low point; trusted less than this administration would be even more shockingly bad.
♥ 1
"To Apply Machine Learning Responsibly, We Use It in Moderation"
This is a great @timesopen piece from @msalganik and @RobinCRLee about what they found when they went checking for algorithmic bias in the @nytimes comment moderation system.
open.nytimes.com/to-apply-machi…
♥ 15↺ 9
Replying to @v0max and @rigow, @Moonalice, @suyashcjoshi
I hold the slightly different position that this change would not significantly impact their revenue (in the way it would eg. for publishers) but it would impact their empire-building.
So I blame greed more than cowardice, but you know ¯\_(ツ)_/¯
♥ 1
Replying to @rigow
Oui. C’est du NextJS, sortir les données c’est crade au mieux, et de toute façon leur source c’est OpenCOVID…
Any mask recommendations for the more nasally developed amongst us? Every reusable mask I've tried so far you could fly a smallish airplane down the outside of one nostril and up the other, I'm not sure it's keeping much in the way of viruses out.
♥ 4
Also, if I wanted to hang stuff off my ears I'd be wearing créoles more often. Why not velcro around the back of the neck?
Replying to @cynddl
Yes, disposable masks are like that and they're OK! I have tried sewing many times and, well, it's probably better if I give someone some money 🤭