Startup offers landlords continuous, deep surveillance of tenants' social media
boingboing.net
boingboing.net
China is already trialling a 'Social Credit' system that aims to achieve similar. The only difference between China and the West is that as an authoritarian state they can be more brazen about it.
Sadly, it being Britain, there will always be a fair share of people willing to go along with it because they 'probably know what they're doing' and the 'innocent have nothing to hide.'
I recall a startup nonprofit that gave loans to people in third world countries. The main problem was detecting fraud, and the only tool they had was checking social media profiles. Clearly this is a good thing if it maximizes the number of people who get loans. Even if a few good people get accidentally filtered, there is a net benefit. And yet this was really controversial in the HN comments.
There is also a net benefit to having more accurate predictions on who will make a good tenant. Rent is the largest cost of living for the poor in many places, and anything that helps reduce that cost is a net benefit.
Yet people react based on emotions. If a single false positive out of millions causes someone to be worse off, it's considered unfair and bad. Even if the net utility is positive and more people benefit than are hurt. It also rests on some underlying assumption that the existing system is perfectly fair, which is obviously false. Human's are way more biased at judging than algorithms. Bad loans and bad tenants have to be paid for by everyone else, which is hardly fair either.
Social Credit has nothing to do with it. The goal of social credit is to discourage dissent. Whereas credit rating agencies don't care what your political opinions are, just how likely you are to pay your bills.
The problem is that a system like this would strongly encourage all people, good or bad, to avoid doing things that the system could object to. It's not enough to count the actual false positives, you have to take this loss of freedom into account as well. I don't think it's a good idea to have people self-censor because they worry how a surveillance algorithm will react to what they do. (The same way that you think twice, nowadays, about using problematic keywords or expressing a dangerously controversial opinion in a text message to a friend.)
I could imagine myself falling foul of this because I produce music. It wouldn't take long on my facebook page to see that, so if a landlord could mine that information then I'd be unlikely to get a tenancy based on my potential to be a nuisance to my neighbours.
That's how car insurance is starting to work. You can get a thing put in your car that will monitor your driving, and in exchange you get lower rates.
Perhaps you could opt into the social media scanner and get lower interest rates. If you don't, you just get grouped with all the other people who voluntarily opted out, and your rates will probably not be any higher than before.
And in any case, landlords and employers already look at social media profiles. It's not like that is going to go away if you ban this, but an algorithm has a chance to be more fair and unbiased at it than a human.
> Social Credit has nothing to do with it. The goal of social credit is to discourage dissent.
That's the law of unintended consequences though. A well-intentioned system designed to determine who is and isn't a good tenant ends up having a chilling effect on online conversation as a whole because people are worried their prospective landlord, business partner, romantic partner won't want to deal with a Tory / Labour voter / UKIP voter / Trump supporter delete as appropriate.
As Schneier points out in his book, it's the same technology used by Disneyland to match people's to their photos when they've been on Big Thunder Mountain as is used to pick out dissidents at political demos in Russia.
Inherent in cost-benefit calculations is a tendency to overestimate short term individual benefits and underestimate long term social harms.
Rent levels have very little to do with actual cost, it's a market-clearing price, only affected by the availability of alternatives.
And what happens to the people who are banned by the system? They don't cease to exist.
Rentiers used to overtly discriminate against people by ethnic group until this was banned, and undoubtedly still do to the extent that they can get away with it. If you algorithm disproportionately rejects members of ethnic minorities, is that even legal?
But that's exactly my point. Humans are terrible and biased. The algorithm isn't racist, it just cares about providing the most accurate prediction possible. Leaving such decisions up to humans is horribly unfair and should be banned, not the algorithms.
>And what happens to the people who are banned by the system? They don't cease to exist.
Is that not already the case with credit ratings and applications and stuff? I'm only arguing that the existing system should be made more efficient.
>Rent levels have very little to do with actual cost, it's a market-clearing price, only affected by the availability of alternatives.
Rent level must be influenced by the number of bad tenants. Evicting people is difficult, and bad tenants can cause a lot of pain to the landlord. Which is why they care so much about using systems like this in the first place.
I don't know, but AirBnB might.
They had some cases where they had no idea why the algorithm rejected some loan applications and then analysed the logs to find out why. It turned out that there was a correlation between specific system fonts and defaulting ob loans.
Giving power over people to opaque algorithms and hustler start-ups is foolish.
It's not that I agree with this application, but it's inevitable that we will be judged by algorithms and we need to be sure they don't act in absurd ways. Public outcry should keep them in check.
The second is that unfairness isn't a problem. Whatever has the most utility should be preferred. If one person is worse off and 99 are better off, it's a net good.
Most people would also say that foregoing the benefits of using Facebook, Twitter, LinkedIn, WhatsApp, Snapchat, Wikipedia, Medium, TripAdvisor et al is an unreasonably large compromise for not having employers, landlords, business contacts, potential romantic interests not running machine psychoanalysis on your social accounts.
Since the people working to put this together clearly don't have the ethical sense to not-make this, at least they might think twice if there's enough understanding that producing flat-out evil might hurt their career prospects.
The background check company got back to me with a list of social media account handles they had found, some were mind, some weren't. They told me that they would be submitting this list to my employer stating that I lied about my social media accounts. I responded by telling them that if they did they would be submitting incorrect information about me and I could take action using the Data Protection Act.
I expected to fail the check, but somehow I passed and got the job.
The takeaway from this is that so long as you refuse to discuss your social media accounts and don't use your real picture as a profile picture, don't have a really unusual name and keep everything locked down and private then how can they prove otherwise?
So lets just say I tell them that I have no social media presence and they dig up some accounts and tell the landlord I lied. Well, they have to justify their claim, they have to be able to conclusively prove it is mine. If they get anything wrong then all it takes is a complaint to the Information Commissioner stating this company is holding, processing and passing on inaccurate data about me to give them a little shock. A few too many complains would then probably trigger a severe audit by the commissioner.
Looking at it from a wider perspective. How are they processing and protecting information in general. How do they protect all the data they harvest from your social accounts? Are staff able to share the info for fun? is there access logging? How secure are their PCs? What do they do with your data after your tenancy ends? This is the sort of stuff a business like this cant afford to make mistakes with. If they do and people complain in big enough numbers it could negatively impact their business.
It's also worth noting that where religion, sexuality or health data is involved things get a little bit more serious as regards the DPA.
If I were dealing this company I would be quite happy to act in manner which would let them dig their own grave even if it impacted me.
You're cracking me up. You can not be pro-technology and pro-privacy at the same time - these things are mutually exclusive almost by definition. Pick a side already and stop whining. (Aaand here's a tip: the winning side in this will always be technology. Yes, that very same technology that you help create on a daily basis. )
First, the obvious; Public Key Cryptography, HTTPs, Digital Encryption Algorithms and their implementations are all TECHNOLOGY and there is a non-trivial number of us HN Monkeys involved in building them as well (I wish I was one, not just yet).
Next. I get your concern. I really do, and I share it. The problem you're referring to is often incorrectly attributed to the proliferation of technology alone - but I think it's a problem of the free market. It's a problem of the commoditisation of Human Attention and Human Data in the quest for ever higher, ever better profits. This is not a new problem, rapacious and horrible entities have been born out of this. My blood boils when I see ads for fertiliser from DuPont on Indian TV. Yes, they're now advertising again as if Bhopal never happened. But hey, profits.
So your anger is misdirected. Please use it as fuel instead to try and build an alternate system. To try and think of ways in which the innate, inescapable human need for increasing individual rewards (today these rewards are expressed in the currencies of Money and Political Power), without compromising the needs of - well - everyone else. We need more thinkers and thinking. It's what humans do BEST.
Hoooooooorse shit.
Actually, probably better not. Escalating mutual destruction scenarios are not a useful long-term strategy.
There is other option: leave. Prague and other cities have many foreigners from UK.
There is already a reciprocal agreement for healthcare and no reason to suppose it would end. Don't forget there are many Spanish people living and working in the UK right now.
So given e.g. https://news.ycombinator.com/item?id=11876453 , what happens?
In the case of Spain and the UK there are lots of reasons to maintain a good working relationship.
Of course it may not work out that way but, going back to the original point, anyone who left the UK 15 years ago had to be aware there may be problems further down the line?
It could be a huge source of revenue for them, or they could decide this would damaged confidence in their service and discourage (my instinct says the latter).
I've also got real social media accounts that I actually use and that are under a fake name and I regularly verify that obvious google searches don't find them.
You can use google alerts to automate this.
Or they find some objectionable friend in his list, by random chance.
Really it's an essay about good password policies, e.g. for Grandma who thinks her cat's name spelled backwards is hard to guess... yet posts all over Facebook about Meowie.
Considering Mark Zuckerberg didn't bother to use LastPass or memorize a simple sort of random string of characters...