Big companies definitely care a lot more about these things than universities or "small-time researchers". I've been in long meetings with corporate lawyers at multiple big companies, and they are a lot more careful than you imply (because they have more to lose). It's always a risk vs reward thing. Things like "copyright doesn't apply here at all. You can use whatever training data you want as long as your model is sufficiently transformative." are not certain until they have been tested in courts, and will be different for each case (it's a lot easier to argue that a simple classifier is transformative than a GAN that might reproduce one of its training inputs, for example). An additional complication is that if you're a company that's operating globally, you have to pay attention to a lot more laws.
Where I work now we work a lot with synthetic training data, because dealing with "real" data is complicated.
One way to protect yourself is buying licenses to images from a service that provides legal indemnification. Part of the reason why stock image websites can earn money is that they provide this, so (at least part of) the risk is with them.
In addition to the concern that copyright doesn't necessarily cover personality rights, you can not assume that just because someone uploaded an image to Flickr as CC-BY that they had the right to do so, in a lot of cases they do not (when people upload images found on the internet). Yes, it's their fault, but in the end you still don't have a valid license for the image.
How on earth are trial versions bad user experience? My #1 wish when I browse an app store would be to look for apps (especially games) that offer trial versions. I don't want to try games that shove ads down your throat or monetize through in-app purchases, I won't pay for a subscription for some mobile game I just play to pass time waiting for something, but I'd like to try the game before I pay for it since lots of games I tried were a lot less fun than I thought from the store page. Trial versions would be perfect.
I'm not sure about "severe consequences". If you compare Switzerland (where, although the government now recommends masks in public transport at least but almost nobody does wear a mask) and Austria (where wearing masks while shopping/in public transport etc was made mandatory when the restrictions were eased), you don't see much of a difference in the amount of new cases. I do believe masks work and I do wear one when I'm going near other people (not when I go for a run) but seeing how few people wear masks here and it seemingly not making a difference does make me wonder how essential they are, and it makes me doubt people blaming the severity of the outbreak in the US on the confused mask messaging.
If it was about being able to spam, they'd just silently strip the + part. I think it's just ignorance ... in my own mailserver I have a custom separator that usually doesn't trip validation functions.
> These deaths could represent misclassified COVID-19 deaths, or potentially could be indirectly related to COVID-19 (e.g., deaths from other causes occurring in the context of health care shortages or overburdened health care systems).
That is not saying "those deaths were not caused by COVID-19". They explicitly mention that these could be misclassified.
The article seems only to focus on what happens if you lose your ways to authenticate, but another possibility is getting caught in some weird ban wave like spamming emotes on a youtube stream when the streamer asks you to (https://9to5google.com/2019/11/09/google-account-bans-youtub...) (most of these bans seem to have been reversed, but I don't know if that would have happened without the publicity that came from a popular youtuber calling out google for banning his fans ...)
I doubt that you can measure this already, it has only been two days (also when the lockdown started it took over a week until the effect was visible). That's why at least some countries want to take it slow, after each small step you need at least two weeks to judge the effects.
That's kind of incompatible with the critique 2 posts up. They asked if we could have "false positive" deaths because dead people with covid-like symptoms are labeled as covid deaths even when they are not tested (e.g. after dying at home). The people on ventilators were ones in hospitals that were treated as covid patients, those were likely to be tested - and the problem that people on ventilators have a low chance to make it is specific to covid (which is why we are seeing this, ventilators have existed for a long time, usually they don't kill people which is why it became standard operating procedure, unfortunately with covid they seem to trigger something bad ...)
Lockdown until you can get back to containment (drive infection numbers so low that you have the resources to trace/test contacts and send possible infections into quarantine).
"Everyone becomes immune" will take longer than vaccine if you don't want to overload your hospitals, and bets on long-term immunity which we don't know yet.
I disagree that a flat case count is positive. That strategy is essentially a bet on others failing to contain the virus. If you assume 90% of infections go undetected, you're looking at 2+ years until you reach 60% of immune people (10M Swedes, ~4300 new cases per week), so the time frame is probably dictated by the availability of a vaccine which - I hope - should be available sooner.
If other countries do succeed in containing the virus, that means travel in and out of Sweden will have to be restricted in that time frame, and death rates will have been higher (and not just a deferral effect).
Since some countries apparently have been succeeding in containing the virus, I don't think that is a good bet to make.
If you look at [3], you'll see that it has been simplified and "manager" isn't required anymore - if you are flexible in your working hours and earn 120k+ (which isn't that much - if you work in tech or at a bank you'll likely earn more) the company can agree with you to not record working hours (there's still legal limits to how much you can work, of course) - this is relatively new, since 2016
Putting Switzerland and Sweden in the same bucket here isn't fair, IMO. In Switzerland, most things were closed down like in our neighbor countries (some things are reopening now, but still very limited). A difference was that while other countries said no groups of >2 people, in Switzerland the rule was no groups of >5 people, so in that sense it was more lax, but it's far more strict than Sweden's policy.
If you don't trust the app to not localize/track you, you also don't want to give it BT permissions. BT on its own is enough to localize/track you pretty well in cities, which was one of the reasons people didn't like the idea of a centralized database of BT contacts ...
> the transmission rate dropped again, to the point where any single case led to less than one more. Once that happens, an epidemic dies.
Note that this can take a while. What the article suggests - getting back to the containment phase - is the stated goal of some countries. However, to get there quick you need very drastic measures. With the measures we adopted, it will take a bit of time - I read one estimate that the reproduction number in Switzerland has been around 0.59 since the lockdown was enacted, so you'd expect cases to drop by a factor of this per generation - it still takes weeks/months to go from 1000+ cases to below 100 (where they estimate they have enough resources for contact tracing).
Germany/Austria/Switzerland have similar programs (called Kurzarbeit) which are for situations like these (where companies don't have enough work at the moment but are believed to recover), https://en.wikipedia.org/wiki/Short-time_working - what was done for the covid crisis here (Switzerland) was to simplify the application process for it.
There will be patches. Ubuntu 18.04 LTS shipped with 2.7 and is supported until 2023. RHEL 7 shipped with 2.7 and is supported until 2024. What will be interesting to see is if there will be a "standard" fork that everybody pulls patches from, or if every vendor patches himself ...
Where did I say the primary purpose was to prevent infecting yourself?
It doesn't help if they reduce transmission, if they increase the risk of infection at a greater rate.
"It never lands on that surface to begin with." is a flawed assumption. You see people taking off/adjusting masks all the time because someone on the phone doesn't understand them, they want to eat something, etc ... or they retie the scarf they use as a mask and don't care which side is the "outside" ... and then go on to touch stuff in the store, and then again their masks ...
Like I said, I used to think it was a good idea, and now I've seen lots of people using the masks.
The goal is still to get back to containment, from yesterday's press conference: "Koch: Das ist genau unser Ziel. Nämlich eine Situation mit einem so geringen Infektionsrisiko, dass man alle Kontaktketten zurückverfolgen kann. Dann wäre es auch möglich, ein etwas normaleres Leben zu führen. Es ist extrem schwierig eine zeitliche Voraussage zu treffen. Der Trend ist im Moment positiv und wenn wir unter 100 Fällen am Tag sind, ist das Contact-Tracing wieder möglich. Das kann in ein paar Wochen aber auch in ein paar Monaten sein."
Re-opening some things is not a contradiction, as long as you keep the reproduction number <1 you are on a path so that you can contain things. Stricter measures mean you hit them earlier, laxer measures means later but you might not damage the economy as much ... but in any case they said if it flares up again the strict measures come back.
Note that this "opening up" on Monday includes very limited things - stores (that have been open all along since they sell food etc as well) may sell more products, hardware stores (that afaik did not close down in the US) reopen, garden centers, and some services like barbers and that's it ... not much a difference IMO, more like a test drive. The one on May 11 is more open and not entirely uncontroversial (the parts of Switzerland that have been hit harder are not exactly happy that schools are supposed to open - and I have to say "kids aren't likely to transmit covid-19 so opening schools is fine" at the same time as "grandparents shouldn't watch kids because they are at risk" doesn't seem logical to me).
They do (otherwise your curve would never end without measures that reduce the reproduction number to below 1) and it's not (only) my calculations, these things (at least in Germany) have been published a short time after the lockdown was enacted (eg https://www.dgepi.de/assets/Stellungnahmen/Stellungnahme2020... - in German unfortunately, number of intensive care beds in Germany was around 30000 at that point so even in the "best case" scenario they look at - 2% of patients need intensive care for 10 days - a curve that does not bust this number extends well beyond the graph which is one year, and not every country has as many ICU beds per resident as Germany has). The "strict measures until we can get back to containment" was also stated recently as the goal by the Swiss authorities ...
Do you have a source for "It's a fact that wearing masks is slowing down the spread by as much as 80%."? There's studies that measure the amount of droplets/aerosol that gets through but I haven't seen any studies about how it affects infection rates.
I initially, when countries here (Europe) started recommending masks, thought that was a good idea, but having seen now how people use masks I definitely know where the doctors that said "people are going to touch their face more due to masks and that's the last thing we want" were coming from. So, I am not sure any more that earlier mask usage would have helped.
That "flattening the curve" isn't going to cut it was obvious to people that thought it through from pretty early on (if you look at the capacity of the healthcare system of your country, and look at how broad the curve needs to be, you end up with "years" for any country I looked at). So, the game plan is to hold out until a vaccine is found and tested. Putting the whole population under heavy measures for that long isn't going to work, so you impose very restrictive measures until containment (tracing infected cases and quarantining those that might have been exposed) is feasible again, all the while taking care that you don't allow big explosions of infections (like they had with one patient in South Korea).
Camera quality is the easy problem. The hard problems are getting camera latency down and dealing with depth of field and all the other things you lose when you replace the incoming light field with camera images.