Facial recognition rolled out by London police sparking human rights concerns
independent.co.uk
independent.co.uk
Add ALPR camera data, and a public records graph, and using those selectors against other areas means you can narrow a list of suspects fairly quickly.
Arguably, surveillance merely polarizes crime instead of reducing it, where if good people are too worried about the consequences of defending themselves, it creates an effective %100 success rate for committing crimes and an insignificant chance of eventual charges - there is little disincentive to to criminals becoming super aggressive.
London is already the most surveiled city in the western world, and yet still notorious for violent street crime. Adding facial recognition to the current regime of cell towers, ALPR, cameras at every corner, cab cameras, and public transit cameras, is meaningless.
What is most likely is that facial recognition will be used the way it is used everywhere else in the world - for political pacification, not crime reduction.
Which I find fairly strange all things considered - the crime index (https://www.numbeo.com/crime/rankings.jsp - admittedly this might not be the most accurate source) shows quite a gap between other similar cities like New York - a fairly common comparison. As a resident of a few years now, I definitely wouldn't say I felt London to have that kind of feel, if anything violent crime has the effect of being shocking because it is so uncommon.
I find it quite disheartening to see London's relatively soft-touch approach lose out over "automated" policing, there's quite a bit of grumbling lately over the reduction in actual people on the streets, mainly over police cuts.
There's nowhere I know of in London where I wouldn't feel safe even walking on foot, and the worst parts keep getting gentrified.
There is danger, all right, but I think you would probably agree that it's not the obvious, conspicuous risks that lead to violent crime. It's the places that don't seem dangerous where it happens, because those are the places where victims are present and not focusing on their safety.
Incidentally these guys are quite a London thing, so if you are at the traffic lights on your bike/scooter/car and you see two or more scooters with no number plates and pillion passengers then you might want to start saying your prayers.
The former chancellor of the exchequer had his phone nabbed by the legendary moped gang thieves so it is not only in areas that are deprived where crime is a problem. Don't walk and text in Central London, you will only be asking for it.
They wear helmets and balaclavas underneath so I do not see how facial recognition can help. Only arming the public with fully automatic assault weapons can deter these ultra-violent thugs who happen to normally be only 15 years old.
London's murder rate recently topped New York's [1]. (Granted, it's probably a statistical fluke [2].)
[1] https://www.usatoday.com/story/news/world/2018/04/03/londons...
[2] https://www.economist.com/graphic-detail/2018/04/03/new-york...
Notorious among American right-wingers, perhaps. I haven't heard this or the "no-go areas" guff from anywhere else.
https://www.bbc.co.uk/news/uk-25022680
'Almost a million crimes a year are disappearing from official figures as chief constables attempt to meet targets, a study by the police watchdog has disclosed.'
https://www.telegraph.co.uk/news/uknews/crime/11236805/A-mil...
As a resident, I have to say I'm pretty surprised to hear this. Is London that different vs similar sized cities?
You can also have a benevolent dictator. But if you avoid the dictator all together, you remove the incertainty.
With orher signals like the brexit, porn censorship, and cctv every where, this is very disturbing.
But the most important signal is how the population accepts it.
It smells like a disaster in the making.
Brexit was 'No' vote to thought-police, political-corruption, and selective outrage.
It was a 'yes' vote for national sovereignty, and empowered diversity of opinion (where a diversity means to be 'different' compared than the 'rich /powerful/famous' line of vote)
To me the early signs of Dictatorship are:
a) Uniformity of thought (and, therefore, active thought police).
b) The culture of Selective outrage (that then leads to selective law enforcement for your citizens).
c) Paid (and therefore 'fake') victims.
d) The identification if a 'single powerful enemy' that must be attacked preventively (or else everybody will get hurt)
---
So a treatment that prevents Dictatorship, is to ensure that the economical, political and military power of a country (or a union) is well spread, and controlled by local votes (as much as possible), rather than representative vote
It does bring 'less efficiency', but it is that type of entropy (introduced by uncertainty of diverse opinion of larger populous) -- that is necessary to prevent dictatorships.
You only have to look at the widespread increase in CCTV use in Britain. Now when you get mugged in London the police can get you a little video of your hooded attackers for posterity. Except they only have time to go through the recordings if you got murdered, otherwise you will just get a crime number.
1) It clearly does not work (as opposed to "it has a low accuracy" which just means they will have to filter manually) and
2) They are pursuing a goal of policing with fewer police officers and
3) Having a working solution would mean they could police with fewer police officers (cameras can't detain suspects or intervene when problems arise, AND police officers don't carry a mental database of all suspects like a computer could have, so having cameras does not replace having police officers).
Personally I've had 3 scooter thefts on the past year, including a bike jacking at a red light - more a robbery than theft.
All these criminals wear balaclavas even if they're not wearing helmets, and they're used to using countermeasures against CCTV, so facial recognition isn't likely to make a big difference to them.
There is a difference between policing and surveillance. Confusing the two is a dangerous - and false - assumption.
The UK has until recently been trying to follow a policy of reduced public spending and reduced government borrowing. It is natural to assume that any new technology is only being used on this basis.
That doesn’t mean this perception is correct, and the current state of the UK looks to me like they’re trying to promise all things to all people, failing to do almost anything, and increasing annoying everyone as a result — so there isn’t any reason to assume logical coherence either.
Um, what? [1]
[1] https://en.m.wikipedia.org/wiki/Budget_of_the_United_Kingdom
Your link just explains that they’re trying to reduce the gap between tax income and expenditure, but the absolute level of expenditure (my link) has gone up every year (in real terms too)!
Talking about the specifics however, the number of police officers in the UK have been reduced by around 19 thousand. Since that only leaves 126k I hope you will find that significant. I find the continued expenditure on unproven gadgets rather distasteful against this backdrop.
https://fullfact.org/crime/police-officer-numbers-have-falle...
When you point accuracy that poor at a sample so large the number of false positives is so colossally large that the resources to 'filter manually' are large, meanwhile a free- country is branding huge numbers of innocent people as possible suspects. So no it doesn't work. For it to work in a free-country we should be looking at five-nine accuracy at least.
> 3) Having a working solution would mean they could police with fewer police officers
But they don't have one...
The really worrying thing for me is they use the excuse that they are looking for > "potential terrorist targets"
and then say
> "poor quality images" supplied by agencies including Uefa
UEFA is the governing body of the European football championship, now obviously famed for fighting terrorism. Terrorism is the excuse they love to use to take away civil liberties.
It continues to worry me
> over 450 arrests
and then discusses 2 convictions. It is a very worrying world we are slipping into where getting lots of arrests counts as a result
> no-one had been arrested after an incorrect match
This is not for the police to decide, it is for the courts. Hi-five 450 and 2 convictions of note
> The technology has also helped identify vulnerable people in times of crisis.
So it's already being used more generally than arresting terrorists, and football hooligans! It crept up on us over the course of one article!
[0] https://www.telegraph.co.uk/news/2018/05/05/police-defend-fa...
I suppose people are expecting it to work 100% perfectly at its debut...like everything else that is brand new.
"Our facial recognition has over 90% accuracy" from the prosecution's expert can easily become "clearly guilty" in the jury room.
I found this article helpful to understand why 90% accurate isn't as good as it sounds: https://www.badscience.net/2009/02/datamining-would-be-lovel...
He goes I to more detail in his book, bad science with a more detailed explanation. It fools everyone because it is counterintuitive.
Most juries will have around half, perhaps more, who've done nothing at all mathematical since school 10-45 years previously and think percentages are in the difficult part of maths (ie it's more than basic arithmetic). HN is going to be very unrepresentative for numeracy. :)
You want them to achieve "beyond reasonable doubt" conclusions from probabilities, percentages, false positive rates and perhaps standard deviations, well you may as well be making your case in French or Ancient Greek that they can pick a few part-understood bits from.
Or hope for an eloquent statistician on the jury!
Often these numbers are given as percentages, and he says that most people (even the doctors and nurses administering the tests) don't really understand them, and that we should use natural numbers (as Ben Goldacre does in the example you provide).
He uses examples from cancer screening and HIV testing.
This. You will have a hard time contending with pretty graphs, confusing numbers and a guy with an impressive title/badge/degree explaining why you're guilty.
FBI Testimony on Microscopic Hair Analysis Contained Errors in at least 90% of Cases in Ongoing Review
https://www.documentcloud.org/documents/1996772-joint-press-...
Ditto bite-mark analysis:
https://phys.org/news/2015-05-fbi-hair-analysis-problems-rev...
Yet, this kind of "expert" testimony was accepted for years and years. People went to jail on the basis of flim-flam.
My point is that you get some bad evidence, some bad eye witnesses or profiles and you get convicted, especially if the judge or jury gets tricked by the media in general on how good the forensic science actually is.
That's not a joke.
In this case it's probably the same people who like the idea of USPS photographing their mail so they can see what they're getting, not realizing they're actually being sold surveillance as a survice. That is, they're happy about it.
I'm more realistic. That scares a lot of people.
[ed: https://web.archive.org/web/20120229200201/http://internetey... ]
You are unlikely to arrest someone based on a mechanical turk results, since unsurprisingly you can look at photos yourself to double check.
It's also pretty common to get data rated by multiple people to cross-check how certain they are, and to keep track of individuals' accuracy and exclude inaccurate raters from your tasks.
2) Surveillance as a Service? That's not a good sign.
If you were part of these organisations you were labelled a 'domestic extremist'.
It is already happening. Has been for a long time.
Police still fight wearing and using body cams. Governments are becoming more obstreperous, recalcitrant, and outright -- and arguable, illegally -- uncooperative with respect to Freedom of Information Act and other like document production procedures and requests.
It seems, the powers that want this, want no aspect of this "transparency" for themselves.
I'd like to see a private initiative and cameras take to tracking the elite and the "nothing to hide" people.
Their actions and behavior already tell me, that they have plenty to hide.
Let's see them "live their everyday life", as they've known it, in the face of relentless tracking and "accountability".
And if the answer to this question is yes, then where does it end? Or will democracies invariably end up sacrificing any and all freedoms and rights of privacy, that can be sacrificed, in the name of security?
GDPR is interpreted and enforced at the national level. I'm going to go ahead and assume the British data regulator won't rule against its own government.
1. A Country that failed to upgrade it's NHS systems from Windows XP after 10+ billion pounds of investment and subsequently got hacked, leading to many deaths - scares me. The data they are collecting is very intimate. If it was leaked, somebody could track persons of interest (government officials, politicians, journalists, undercover police, etc), tell whether somebody is having an affair, make correlations between friends groups - tonnes of data. They could potentially figure out where you work, where you go for lunch, who you meet, who you phone - it's just too much.
2. Today it's about catching "known" criminals, tomorrow it's about predicting future criminals and sending them to state "rehabilitation" camps (just look at China), selling you data to the highest bidder (it's happened on the Council level in the past) or some other ridiculous use. Blocking websites on the internet used to be about removing exploitation of young persons, now it's used to block pirating of copyright material.
3. A right to be forgotten - if you've done nothing wrong, you shouldn't be tracked. I can understand the need to hold this sort of data for a short period of time, but after this it should be deleted. I think privacy is a fundamental freedom, along with freedom of speech, freedom to travel, etc.
4. The accuracy of facial detection, even if it's 99.99% for 70 million people, is still 7000 mistakes (at a massive simplification). It's actually worse than that, both the real accuracy of face detection and the error rate for specific people. It's also "easy" to trick. When it works it works well, but what about the person who is incorrectly detained, potentially injured for a crime they never committed? And that's not even taking into consideration that a lot of these face detection algorithms tend to be subtly racist (over trained on locals, badly trained on foreign persons).
5. Lack of oversight for such a powerful technology - to collect data on this magnitude is absolutely insane and needs multiple groups of people to keep it's use in check. Not just government based but also separate non-government groups which receive funding elsewhere.
That all said, I can understand the clear benefit - remove the human element from tracking these people. But with any powerful technology we must always ask - at what cost? I think this inflicts too much on a persons basic freedoms.
Even human beings are not always able to identify illegal activities, let alone AI.
Any given AI won’t spot general unlawfulness at this point, but they are getting designed to spot specific laws getting broken.
Perhaps I could have phrased my “in general” better.
I don't drive, so no :)
More to the point, figuring out the speed of a moving object is not an AI task- and I don't mean that in the sense of "if it works it's not AI". I mean, really, it's not something AI was ever interested in, presumably because it's not a particularly complicated calculation, given the right equipment.
Generally, AI is interested in problems that demand, how can I put it, unorthodox solutions. Or just very tricky ones.
So the kind of thing I thought you meant was identifying, say, burglars or muggers, from video feeds etc. That sort of thing is not possible yet, certainly not outside controlled conditions ("in the lab").
Reading the number plate, on the other hand…
I would be surprised if the AI in any worthwhile self driving car couldn’t detect 90% of categories of unlawful road use.
Detecting assaults may be computationlly unreasonable at this point, but there is work on generating 3D meshes which map to all human bodies in a scene, so it’s not unreasonable to draw a line from one to the other. Identifying that a theft has occurred, however, probably can’t be done yet outside carefully controlled conditions. Yet.
(I have been given one speeding ticket, but in error because I had sold the car before the incident).
Yep, that's a honest-to-god AI task :)
>> Detecting assaults may be computationlly unreasonable at this point, but there is work on generating 3D meshes which map to all human bodies in a scene, so it’s not unreasonable to draw a line from one to the other.
I think the closest analogy is pose estimation, where there's quite a bit of work (in particular, I think there's a lot of interest in learning to identify body postures that can lead to a fall, in order to reduce injuries to older people). I don't remember seeing work on identifying criminal intent in particular, though.
My intuition is that it's more than a matter of computational resources and will require some algorithmic advances. But, you never know.
>> (I have been given one speeding ticket, but in error because I had sold the car before the incident).
Automation, eh? :)
Yup. No idea how it went wrong given I only owned it for only a few weeks specially so I could sell it on behalf of my partner after she accidentally moved to America (I have a complicated reality [1]), and the DVLA sent me acknowledgment, about nine months before I got the letter from the police, that I had sold it.
[1] “And that’s how I found out that Michael Jackson works for the USAF.”
There is very little understanding in the UK, and in England in particular, that it is dangerous to allow a state to become too powerful, for its police force to be so effective that it precludes the possibility of the public removing a government that is not serving its people.
Privacy is a human right.
do you not feel comfortable revealing your identity alongside your words?
now you have a taste of why privacy is important.