Or even simpler – we identify a guy in a marginal constituency who interacts with a lot of other people we have previously idenified as swing voters. We bombard him with propaganda, tailored to his interests using the Facebook information we already hold about him.
So nothing they can't do already. And as for the protest example, nothing they don't already do via facial recognition.
Happy for a split in the thread, although this is a bit crossover.
False choice. You have to be at home (it’s the law) but now your saying you’re happy to let a group of people you know have a track record of manipulation of data a free hand to do what they want with it, regardless of the fact there’s a better, safer and more productive way of doing it.
I'll use the the better, safer and more productive way of doing it, of course.
If the NHSX offering proves popular I'll use that as well. Clearly, there's no point if I'm the only one using it. The govt have all of my medical data, my tax data, my education data, all of my vehicles details. I have to put a number on my car to uniquely identify it.
I’ll use the the better, safer and more productive way of doing it, of course.
No, you won’t we able to unless the UK gov changes their approach.
The issue here isn’t that the government will know where you were a week last Tuesday, it’s that they may (will) use the information they gather to undermine democracy.
How, exactly? Spell it out for us hard of thinking? What will they have that they don’t already have from social media, tax and employment records, NHS records?
I have concerns with the proposal, I can see the arguments for something better, but I’m not seeing an obvious opportunity the clear and present danger that makes me think I wouldn’t download the app.
A lot of you seem to be mixing up your hatred of this wretched government with an over searching conspiracy theory that makes you come across a little bit “tin foil hat”.
The App is being run by the NHS, not the Government.
I know we've all been indoors for a long time, but we're not governed by a load of all knowing, all powerful overlords. Our Public sector is compartmentalised and largely run by civil servants.
"The Government" which seems to be a catch-all title for The State / The Man. Is 26 people working out of Westminster. They are mostly on the same side and mostly loyal to the PM, well, you know until there's a chance to plunge a knife between his shoulder blades and take his job.
The Intelligence Service is a non partisan organisation that tackles National and International Crime and protects the UKs interests both home and abroad. There's about 10k of them who worry about domestic issues, they're split about evenly between Crime and Terrorism. They're about the only organisation (along with GCHQ) who could, in theory hack an NHS app and use it for nefarious means, but they won't pander to the Government of the day.
It's easy the think that someone like Dominic Cummings can log onto his PC in whatever hole he calls home and just have a look at everyone's NHS records, HMRC records, call up live Spy Satellite imaginary etc but it doesn't work like that. There are too many people in the loop that if they say illegal activity like would blow the whistle, either because it suited them personally or because they know it's wrong.
They don't need to anyway, enough people willingly give all sort of information to the social media and search engine providers, it's far cheaper and easier to pay for the information, than steal it.
Except Google & Apple are just building the APIs to provide this functionality in the OS. You’ll still need an app to call those APIs, and G/A have said they’ll only support one app per health authority.
So G/A are allowing access that normal Apps don't get? So whoever does the app gets a monopoly?
It would have saved a lot of time of you'd read the article oob
The govt have all of my medical data, my tax data, my education data, all of my vehicles details. I have to put a number on my car to uniquely identify it.
There are also regulations on how they use that data https://www.gov.uk/data-protection. As yet, they haven't confirmed how their app will comply with those regulations:
https://twitter.com/jimkillock/status/1257649614188142592
So whoever does the app gets a monopoly?
Yes. A monopoly is required because everyone needs to be using the same app. Do you think Apple & Google have worked together on this for shits and giggles?
It would have saved a lot of time of you’d read the article oob
Yes, I apologise for not doing so.
...anyway, now I understand. The option is give a monopoly to the NHS, or give a Monopoly to Apple and Google, and we're just discussing which of those two should get the Monopoly.
Cool.
Well, yes, but one would work, the other won't
The option is give a monopoly to the NHS, or give a Monopoly to Apple and Google, and we’re just discussing which of those two should get the Monopoly.
No, the “NHS” could be the sole user of the joint Apple/Google API for their territory, but has (or rather Cummings&Warner&Warner&co have) chosen to use another solution that means that no one will be using the joint API in England&Wales, despite its obvious benefits. We’re too special.
I’ll use the the better, safer and more productive way of doing it, of course
Which is?
(Sorry if I’ve missed it in a fast moving thread)
Can we get an app thread, keep this relevant please mods,
I've looked to see if anyone has announced an alternative app that uses the Apple/Google tools for the UK but haven't had any success. 😕
You’ll still need an app to call those APIs, and G/A have said they’ll only support one app per health authority.
No, the “NHS” could be the sole user of the joint Apple/Google API for their territory, but has (or rather Cummings&Warner and friends have) chosen to use another solution that means that no one will be using the joint API in England&Wales.
I can't reconcile these two statements.
Can someone post a link to the register article, I've gone back 5 pages and not found it.
Is this it, doesn't seen to answer much:
https://www.theregister.co.uk/2020/05/04/uk_covid_app_human_rights_parliament/
I’ll use the the better, safer and more productive way of doing it, of course
Which is?
Beats me, I took the post I was replying to at face value. I'm gagging to know. Hopefully someone will post the register article.
The second statement you've quoted I don't think is strictly correct. Somone could (maybe has) develop an alternative, but if it's not feeding in to the rest of the test/track/trace architecture is it of much use?
I'd argue it would be.
Perhaps NHS Scotland will do it? They've been a bit cagey...
The second statement you’ve quoted I don’t think is strictly correct.
I've read the three article register articles now and I think both statements are wrong. Or certainly aren't supported by the articles.
Anyway, my curiosity is satisfied I know what's going on now, thanks Del.
excess deaths
Correct - all-cause mortality. I'd plotted cumulative deaths - when I plot weekly the peaks are a bit closer together and they do appear to be post the peak for weekly incidence now. Not a great plot 🙁
my curiosity is satisfied I know what’s going on now
I’m glad you think you do.
Not a great plot
Still some seed for hope there though. I was worried that the “outside hospitals” peak might still be some way off… looks like it may lagging less than a week behind hospital deaths.
Regional variations look scary, but I don’t think they are … just a tad behind … of course those are regions that might have got off far more lightly if we’d taken measures only slightly sooner.
The option is give a monopoly to the NHS, or give a Monopoly to Apple and Google, and we’re just discussing which of those two should get the Monopoly.
Google and Apple have created a joint framework that other parties can use to develop contact tracing apps. They are not developing the apps themselves. Google and Apple have additionally said there can only be one official contact tracing app per region/territory, to avoid making contact tracing ineffective because of multiple competing apps.
The Google/Apple recommended approach uses a decentralised architecture which is less risky from a privacy perspective, and avoids the technical problems of requiring users to have the app running all the time in the foreground on an unlocked phone. The UK has decided not to follow this approach and is using a centralised reporting system which is unlikely to work because of the aforementioned technical limitations of the way smartphone security works.
Unlike the UK, some other countries have decided that developing their own apps based on the Google/Apple APIs will be more effective than trying to design their own architecture which doesn't work properly with smartphones.
Kcr has summed it up.
There's another article here
Sorry - thought the register covered the 'single use' aspect, but this is where is read about that.
Regional variations look scary, but I don’t think they are … just a tad behind
Surprisingly, at the regional level (and below this) thinks tend to look very uniform when you rebase to the same point (number of cumulative deaths is my rebase). The variability in cumulative cases and cumulative deaths is actually small. That says that there is a common transmission path. London is in the lead, South West is lagging most. Generally the further from London, the longer the lag. 2.5 - 3 fold increase in deaths/week at peak.
We got 'lucky' in that lockdown occurred before both the good weather and Easter.
when you rebase to the same point
Yes, that’s what I mean by looking scary… they look slower to reach peak… but are really just behind.
Kcr has summed it up.
Summed it up perfectly. But so have others already. OOB is just choosing not to understand (I hope).
Yes, that’s what I mean by looking scary… they look slower to reach peak… but are really just behind.
For your delight, I plotted ratio on a log scale - so you really can just shift each region along till it meets London. The notable thing is that London increases faster than the others, which are basically the same.
AIUI the Australian app is using a similar model to the NHS app, ie not (currently) using the Google/Apple APIs so needs the app open, uploads to a database on a positive tests and their centralised data is being processed by AWS which is a whole extra layer of privacy confabulation to worry about. They're a few weeks ahead of us, so what are their results looking like at the moment?
The App is being run by the NHS, not the Government.
No it's not. It's being run by Faculty and the NCSC and all the information is being stored on the NCSC servers with the NHS having access to the data to do the bespoke warnings etc they are hoping to be able to do.
The information is then being stored, indefinitely, for 'research purposes' with no limits or definitions as to what that research may be and with the ex-staff of Cambridge Analytica on the steering group with what to do with said data in the months and years to come.
I quote - from government sources - the following:
---
Levy also glossed over the fact that as soon as someone agrees to share their information with UK government – by claiming to feel unwell and hitting a big green button – 28 days of data from the app is given to a central server from where it can never be recovered. That data, featuring all the unique IDs you've encountered in that period and when and how far apart you were, becomes the property of NCSC – as its chief exec Matthew Gould was forced to admit to MPs on Monday. Gould also admitted that the data will not be deleted, UK citizens will not have the right to demand it is deleted, and it can or will be used for “research” in future.
---
It would be so easy for the government to say - 'ok when this shit is over it all gets deleted' but no, it's being held on non-NHS servers for 100 years or more and with people with expertise in extracting useful electoral data from large datasets potentially having access to it.
I may be wrong but the app seems open to abuse regarding false positives/trolling?
If it is not identifying the user, then you are self-certifying that you have tested positive for COVID-19 and there would be no way to verify that.
So if someone is a bit hypochondriac or malicious (i.e. deliberately gets on a crowded train to be close enough to people to register as a possible contamination contact) then they signify in that app that they are "positive", lots of people have to unnecessarily self-isolate for 14 days.
Or do you have to prove your identity and positive test result to be able to notify as positive in the app?
I may be wrong but the app seems open to abuse regarding false positives/trolling?
Yes, this (and all the other privacy/security issues) are why it is really hard to build something like this and get it right. You can't just rush through some shonky app, tell everyone it is their duty to use it, and hope that things turn out OK.
I'm sure that it's possible to flag users who keeping spamming positives and exclude their updates.
Just in case anyone was in doubt as the true nature of the government's apparently personable health secretary.
https://twitter.com/sturdyAlex/status/1257634177643577346
It would be so easy for the government to say – ‘ok when this shit is over it all gets deleted’ but no, it’s being held on non-NHS servers for 100 years or more and with people with expertise in extracting useful electoral data from large datasets potentially having access to it.
That would be some fairly shockingly shit work on the contract. Could lead to the interesting prospect of post pandemic the UK government paying to use it's own data.
Could lead to the interesting prospect of post pandemic the UK government paying to use it’s own data.
But why would they need to use it post-pandemic?
A useful reminder/summary of how we got to where we are… some are likely to forget, as the UK tries to focus on the “successes” of this government during this health emergency:
I’m sure that it’s possible to flag users who keeping spamming positives and exclude their updates.
A bad user only has to notify themselves as positive once.
Yeah we can definitely trust Johnson & co with our data, it's not like they have previous....
https://www.buzzfeed.com/alexspence/boris-johnson-dominic-cummings-voter-data?ref=hpsplash
Yeah we can definitely trust Johnson & co with our data
And trust is the issue here. There may be nothing untoward at all but the fact that important points, such as data retention, the fact it's not really the NHS running the show, where the data will be stored etc has to be dragged out of the government rather than put on the tin so to speak makes people suspicious.
Why not just put it all out there and let people make an informed decision rather than have select committees and journalists drag it out bit by bit?
Is it because they think people won't trust them and therefore decide it's better to go light on the detail or is it because we can't trust them as they have ulterior motives? Either position isn't great.
The Google/Apple recommended approach uses a decentralised architecture which is less risky from a privacy perspective, and avoids the technical problems of requiring users to have the app running all the time in the foreground on an unlocked phone.
More importantly the Apple/Google approach issues a new anonymised ID each day where as the UK approach assigns a single ID 'for life' which is tied back to at least the postcode you input at sign-up and the specific make/model of phone you use.
The A/G approach makes it much much harder to use the data for anything other than contact tracing / warning you if you've been near an infected person.
which is tied back to at least the postcode you input at sign-up and the specific make/model of phone you use
I’ll put money on IP address being captured at sign-up and/or symptom/test submission as well. Unless they explicitly declare otherwise.
The A/G approach makes it much much harder to use the data for anything other than contact tracing / warning you if you’ve been near an infected person.
And that is why it isn’t being used. Or at least it is hard to avoid that suspicion. Why chose to go down a route that clearly will be less complete when it comes to the actual task in hand, hand shaking with nearby devices, other than to avoid the limits on data capture that the big two have put in place to try and encourage as many people to opt in as possible?
So to summarise. We should set aside all the worries that we have over supplying access to our data to the likes of Cummings and co 'because it's what's best for the good of the country'
Should the same discussion not be happening in committee rooms:- 'we should set aside our plans to get access to all this data despite all the things we could do with it in future, because now it's known that Cummings and co are involved people won't take it up. It's what's best for the country'
and to answer 'what data'?
I don't know, I'm not an expert - but as soon as Cummings is near it, even my inexpert senses start tingling. Convince me he has no ulterior motive, because it'll be the first time.
I'll be installing the app as I hope that doing so is the lesser of the evils.
Hate (not a word I use lightly) the government and Cummings but they have us by the balls on this.
Just in case no-one's noticed there's another thread for the app appeared. Should maybe take it over there and free this thread up?
https://singletrackworld.com/forum/topic/the-contact-tracing-app-accuracy/
So it isn't just Scottish Govt senior advisers who don't think it applies to them?
A prime example of 'Do as I say, not as I do'.
Just been reading about Ferguson's involvement in the modelling used to (very wrongly) predict death rates during foot & mouth in 2001.
That model was widely criticised at the time and by the Royal Society in the gov ordered inquiry.
That flawed modelling resulted in 7.7 million farm animals being slaughtered and did massive damage to the rural economy.
Doesn't mean that his/Imperial's models wrong this time.
Probably got caught whilst trialing Doms app.
Thanks for your posts, as always.
Surprisingly, at the regional level (and below this) thinks tend to look very uniform when you rebase to the same point (number of cumulative deaths is my rebase).
Does that mean that the regions where the lockdown occurred earlier in 'their' outbreak still hit the same max R number and the same number of deaths relative to their population? And also that population density seems to have made no difference?
If Regions in the Uk behave the same why don't countries in the world?
The variability in cumulative cases and cumulative deaths is actually small. That says that there is a common transmission path.
Thanks, what is meant by "common transmission path"? (I'm guessing the answer to this answers my third question.)
Just been reading about Ferguson’s involvement in the modelling used to (very wrongly) predict death rates during foot & mouth in 2001.
That model was widely criticised at the time and by the Royal Society in the gov ordered inquiry.
That flawed modelling resulted in 7.7 million farm animals being slaughtered and did massive damage to the rural economy.
Doesn’t mean that his/Imperial’s models wrong this time.
That's not really fair. They didn't know how long FaM was infectious for in 2001. Nobody did, years later on they found out:
https://www.bbc.co.uk/news/science-environment-13299666
Doesn’t mean that his/Imperial’s models wrong this time.
All predictions are wrong, we just don't know which way they're wrong.
All predictions are wrong, we just don’t know which way they’re wrong.
or, as I read recently, all models are wrong but some are useful
all models are wrong but some are useful
Excellent, I'll steal that one.
Hate (not a word I use lightly) the government and Cummings but they have us by the balls on this.
No they don't. They can't compel us to use it and if sufficient numbers won't they will have to adopt something different and better - which apparently exists.
The only good thing you can say about Ferguson is that at least he fell on his sword straight away when he was caught. Beggars belief that he was behaving in a way that undermined the conclusions of his own research.
I did smile at the fact his "married lover" is called Ms Staats...
They can’t compel us to use it and if sufficient numbers won’t they will have to adopt something different and better – which apparently exists.
Or do as they have in Australia, and introduce laws on the use of the data to reassure people and increase adoption. If people refuse to use the app, they have to work out why and sort it.
Agree that it’s easier (and more effective) to just change the whole project to use the Gooogle&Appple approach at the last minute before going countrywide, like the Germans apparently have.
oob - director of the Pirbright Institute for Animal Health said, at the time, that Imperial had produced 'some very seductive graphs' based on 'seriously inconsistent data'.
See current Private Eye, pg 17 'A Sage Fellow?'
at least he fell on his sword straight away when he was caught.
Is that good? His ICL report directly lead to the UK lockdown. If we think his report was pretty much helpful, why would we want him out of the game?
Same with Calderwood.
If we think these people are crap scientists, fine, but if we think they're capable they need to stay in role. Even better maybe the press need to stop following them around.
oob – director of the Pirbright Institute for Animal Health said, at the time, that Imperial had produced ‘some very seductive graphs’ based on ‘seriously inconsistent data’.
See current Private Eye, pg 17 ‘A Sage Fellow?’
Fair enough, thanks for pointing that out.
France going its own way as well:
https://9to5mac.com/2020/05/05/france-issues-misleading-statement/
Is that good?
Yes, because leaving him in place would send the message that it's OK for some people to break the rules. There will be a whole lot of people out there tonight thinking, "well if he thought it was OK, then..."
Similar to Calderwood's case, either Ferguson doesn't really believe the conclusions of his own research, or he does believe it, but thinks it doesn't apply to him. Either way, it fatally undermines his credibility as an adviser in the current situation.
I've not seen my gf for about 5 weeks. Cheers Neil.
Louise is contagious?
Did u mean this. https://www.biorxiv.org/content/10.1101/2020.04.29.069054v1
Interesting May explain why there were less deaths in China than many expected to see
What would be worrying for them.is it this strain backwashes into china (so far it hasnt)
New , more contagious strain is shown in blue
UK death figures causing Johnson to go back into hiding, Raab talking rubbish as usual, whilst showing a graph comparing UK to different countries saying it's wrong to compare different countries
Testing numbers consistently below 100,000 -the 40k in the post ones having vanished , leaving us well below where we need to be to end lockdown
Hancock being a patronising tool in commons, when faced with tough but reasonable questions from an a&e doctor
Patel being clueless on why we didn't quarantine incoming flights
Dep science advisor saying we should have followed WHO /Singapore style advice advice, after all
Witty saying we should have tested more but didn't have capacity (which isn't true, the machines were always there in unis, research institutes & private labs & could be been used sooner)
Senior SAGE member being forced to quit for shagging around (& being exposed by Torygraph who have lionised Johnson's philandering)
Gove hell bent on not extending brexit transition above everything else, while there really is something far more serious to deal with.
It's hard not to be worried about how much worse the government seem determined to make this
Imagine having to resign from a position working for Boris Johnson because you’ve been shagging a married woman, I mean, how unlucky is that?
Little head ruling big head....
If Regions in the Uk behave the same why don’t countries in the world?
They are for the most part. That’s the surprising thing. It’s a common transmissibility. Prior to lockdown there was a very consistent doubling time for cases and deaths. The U.K. was on the wrong side of that distribution. But the distribution was fairly tight.
As for all models being wrong. That’s from George Box, a statistician if note. Models help inform policy. But conclusions should be robust to assumptions. The assumptions in the Imperial model produced a large burden on healthcare for a wide range of plausible assumptions. As did even the simplest back of envelope calculations this time around. Hence the lockdown.
As for my former colleague, I won’t comment. But my former boss was right to hilight that the mistakes fall in the testing arena. PHE will not come out of this well.
My sister is an OT. She put her hand up for supporting during the covid response. She's had some harrowing experiences over the past few weeks. Her account from a private care home that she had to visit last week was more horrific than anything I have seen reported on the news in the UK.
After being exposed herself, she's now isolating awaiting results of a test she had done today. We are keeping our fingers crossed...
There are some unsung heroes out there and she's one of them.
I have only been out 4 times in nearly 7 weeks now and even that was out driving my car to charge the car battery.
I have not eaten any eggs, bread or drink milk for nearly 4 weeks now.
Stay safe everyone. Hopefully this nasty will be over soon.
Not sure I will fly home to the far east this year due to risk ...
If Regions in the Uk behave the same why don’t countries in the world?
They are for the most part. That’s the surprising thing. It’s a common transmissibility. Prior to lockdown there was a very consistent doubling time for cases and deaths. The U.K. was on the wrong side of that distribution. But the distribution was fairly tight.
Many thanks, that's clear.
Prior to lockdown there was a very consistent doubling time for cases and deaths. The U.K. was on the wrong side of that distribution
Can you expand on what you mean here?
Cheers for posting that, I’d not seen anything on the spread of the two different strains for a while.
On the presumption that I have had it, and I’ve definitely had a chest infection that includes, coughing, fatigue, confusion*, temperature, metallic taste, toe rashes, localised aches/pains, sore throat, tight chest, and whatever stuff I’ve since forgotten about. It looks possible that I’ll have had the milder of the two. I first noticed symptoms beginning of late February/March, I want to say March 2nd but that might just have been a Packrafting trip I cancelled.
That would suggest infection mid to late February.
*arguably the most unsettling part
So the new Blue strain - is that what we are really trying to avoid as a “second wave”? Or are we still worrying about a second wave of the original strain and this comes on top?
Theres a lot of implications in the business. Press and media using the words “three weeks” to come out of furlough as if everything will suddenly spring back to normal. Even my own bosses - the very top one being a Trump supporter so I’d debunk him - are spouting about customers being back to thier offices soon. For a bunch of very intelligent people in high positions I can’t help wondering if it’s me... but surely not.
Can you expand on what you mean here?
The coefficient of variation for doubling time was about 30%. From that I created 95% confidence intervals for this growth rate across all countries excluding The U.K. when I estimated U.K. doubling time, it was higher than the upper 95% confidence interval. Hence the rate of increase in the U.K. in early March was significantly higher than the rest of the world mean and we were an outlier.
The doubling time for deaths in the UK was 1.9 days. Globally it was 2.9.
Even my own bosses – the very top one being a Trump supporter so I’d debunk him – are spouting about customers being back to thier offices soon
We are deemed non-essential workers. I’m not lab-based. Our return plans most likely don’t feature going back into an office until about September. We won’t be racing back. And the digital change has been dramatic
I still don’t really know if he should have stood down or not, my own knee jerk reaction is that he should, but anyway… an interesting take on it…
https://twitter.com/botanygeek/status/1257937836243853318?s=21
It does feel like an “easy” story for the Telegraph, when there are “hard” ones that need writing that would implicate their man at No 10.
He was still in the wrong though… can’t see how he could keep his role.
It does feel like an “easy” story for the Telegraph
Of course it is. I bet most of the right wing press couldn't believe their luck when this fell into their laps.
I still don’t really know if he should have stood down or not, my own knee jerk reaction is that he should, but anyway… an interesting take on it…
If I've got it right, he'd already had Covid-19 with a confirmatory test and had decided that he had a level of immunity, so while he may have committed a 'crime' against lockdown regs, in reality he wasn't going to catch or transmit the virus. Not a great look for sure, but I'd rather have competent people engaged in combatting the spread of this thing than sat on the sidelines because of a low/no risk encounter between consenting adults.
Also some classic double standards going on as ever. If this is a resignation offence then so too should Johnson's hand-shaking debacle be. And the rest.