Trump, Brexit, and Cambridge Analytica – not quite the dystopia you're looking for According to a story doing the rounds, psychometric big data pushed Britain into Brexit and Trump on to America. The winning sides adopted a method developed at the University of Cambridge to psychometrically profile people by using publicly available data including Facebook "likes". They used these to create devastatingly …
Had it not been for the "liberal and artistic" it might have confirmed the speech was written by Nick Timothy {caution: Daily Mail).
I was waiting when will the register wade into what is quite clearly an IT topic. It is only one week after the guardian and only one week after MPs have started raising the question.
Congratulations. Tech news indeed.
As far as Analytica and SCL are concerned the interesting statistic is not their claim to "rigging elections the way we like it in 15 odd countries". The interesting statistic is how many of the known ones are now in civil war. The obvious question is also "how many of the ones we do not know about where sh*** have hit the fan have had them intervening in the democratic process as well?".
Never before in human history has it ever been possible to profile so many people based on their activities with digital devices like their mobile phone, browser activity, medical records, school records and any other record that exists in the hands of the military/Govt of a country.
Brexit & Trump is the UK & US equivalent of the Arab spring for the Middle East.
Its amazing what you can do when you have a minimum level of data, but of course the likes of the GCHQ, NSA and others tied up with things like the UK Behavioural Science Unit (Nudge Unit) will always be trying to pre-empt the next terrorist or criminal act.
So as they push your buttons, push their buttons and get them to expose themselves. It might take a few years or decades of provaction, but they will eventually show their hand if you look hard enough.
b1gbr0th3r15h3r3
Regardless of its other uses, some are worried by politicians using psychological profiling.
During the 2012 presidential election, in battleground state Michigan, the Obama team ran slightly different TV ads in different Michigan counties and towns. They did this by working with the cable TV companies to change the ad based on the subscriber's location.
With many people not owning home phones, and unwilling to answer a cell call from a number they don't know, telephone polling can't be trusted anymore.
For the past six months I've working with some very smart friends on a system similar to this article. Using publicly available information, we are working on predicting public trends, and also voter behavior. We got the idea from Asimov's Foundation series and what he called "psychohistory". It uses a modified version of Gregory Bateson's psychological logical levels. Very cool stuff.
> It is only one week after the guardian and only one week after MPs have started raising the question.
And the Reg really downplayed this in its headline when it reported the ICO was investigating Cambrisge Analytica, which struck me as odd at the time. The strange thing about this article is that is says "a data analysis company backed by a Donald Trump-supporting billionaire" but doesn't fucking name him. What the hell? Robert Mercer.
Even for the IT angle, the billionaire in question is worthy of note (he has an interesting IT past). It's like the Reg hasn't even read his bio, yet the Reg is telling us 'there is nothing to see here'. Said billionaire is deliberately and knowingly supporting the spreading of demonstrably false opinions as facts. For gawds sake Reg, doesn't that upset you as journalists?
Private Eye is enjoying record sales.
So by analysing what people say & do, you can work out how they might vote? Not an especially remarkable discovery, I'd have thought. After all, people in N.Ireland have been doing that for decades, based on such traits as how one says the letter "H".
How do they get from there to influencing votes? If I'm a leftie voter I probably prefer to read leftie articles and newspapers. I can't see that they'd have much luck changing my mind with some targetted ads for the Daily Mail (or the reverse).
I can't help but feel there's some confirmation bias here. People want to believe that anyone who voted Trump, or Brexit, must be a gullible fool who is incapable of weighing the issues and making a decision. Pointing to some psychobabble about how it was all down to clever advertising that fooled those gullible idiots helps them avoid accepting the unpalatable reality; that some of those voters might well have voted as they did after a reasoned evaluation of the issues. Not all, perhaps, but enough to make a difference. A bit like mediæval peasants: "I don't know why it happened, it must be the Will of God". A handy excuse to explain something unpleasant.
The psychobabble must be gaining traction. I saw Barbara Streisand was blaming Trump for her weight gain so it seems that minds are going to mush. Although I grant it may not have been a big change for some.
"If I'm a leftie voter I probably prefer to read leftie articles and newspapers."
So, if like me, you read a full spectrum of "news papers" how do they profile that?
I read the Guardian, the Telegraph on a regular basis, I often flick through the Mail of a Saturday morning in my local coffee shop and I've been known to look at the Express and The Sun.
Not sure I'm seeing the problem the same way as others here :-(
It's certainly not (imo) rigging an election, that's done by fiddling with ballots, prevent certain groups from voting, or directly threatening groups to vote a specific way 'or else'
This is targeted advertising, finding a group of people who are likely to be susceptible to a certain message and then giving them that message.
This method is available to every political party and leaning, so all can use it to deliver their message to the groups they think will appreciate it most.
The article infers that it's a cheap way to get a message across, so in fact you could argue that it levels the playing field, it's not just the player with the most advertising pounds that wins any more.
Whatever your personal feelings on both the Trump and Brexit result, surely the smaller parties having access to spread their message as readily as a larger well funded party is a good thing?
>This is targeted advertising, finding a group of people who are likely to be susceptible to a certain message and then giving them that message. This method is available to every political party and leaning, so all can use it to deliver their message to the groups they think will appreciate it most.
Mercer doesn't just operate in advertising, that's the issue. So no, their tactics aren't available to all. And do please note, this Reg article only addressed a small part of the Observer article, so it can't claim to refute it.
The issue in question is 'fake news', but not 'fake news' coming from the established media (for all their faults - shit, we're now in a world where Fox News looks sane in comparison) but from upstarts. This strategy has been used in the Putin government for some time - they aren't trying to get you to believe their version of events, but to be incapable of accepting *any* version, leading to division and paralysis. Just like Trump, the Putin government even told everyone that confusing the hell out of everyone was their aim.
Oh, that reminds me Reg - what happened to that Q&A with Adam Curtis we lead to expect on the Reg? It was never followed up.
It's certainly not (imo) rigging an election, ...or directly threatening groups to vote a specific way 'or else'
Dunno about that.. There was lots of "vote CMIC or Shrillarity willl raise taxes" and "Vote Killary or CMIC will start newklaaaar wars!". Some of that could be called "vote a specific way or else".
Of course, is par for the course for any election anyway.
Except if I remember correctly, Psychohistory only worked when applied to large numbers of people, and only on those who weren't aware of it?
1. It was applied to large number of people - millions of voters. Not so much "swing" ones, but mostly the ones that otherwise did not bother to vote.
2. The actual population it was applied to was not aware it was being targeted. They are still not aware they are being targeted as they are not the person who will read El Graunidad and El Reg.
What this means for democracy (regardless of what the dear professor with the lunatic opinion says) is that democracy is now purely a matter of money. Whoever has the money to pay for the datasources and the compute resource to utilize them has a distinct advantage in the next elections. Neither one comes cheap. You are looking at anything between 5 and 30-40 million "per application".
"What this means for democracy (regardless of what the dear professor with the lunatic opinion says) is that democracy is now purely a matter of money."
But didn't the campaigns that spent significantly less money win in both Brexit and the US Presidential election? While the help may have been donated, Trump spent around US$240m vs Clinton on US$450m (http://www.cnbc.com/2016/11/09/trump-spent-about-half-of-what-clinton-did-on-his-way-to-the-presidency.html)
The spend for Brexit was around £16.4m for leavers vs £15.1m for remain (http://www.independent.co.uk/news/uk/politics/brexit-latest-news-investigation-eu-referendum-vote-leave-britain-stronger-in-europe-campaigns-a7596981.html) and the official campaigns spending £7m each. I would argue that remains platform (i.e. the majority of politicians supporting remain and getting media coverage as the incumbent), the campaigning before the official start by the government (£9m spent on leaflets to every UK household) and the split between the two leave campaigns may alter the effective spending of each side in favour of remain.
Based on buying an election - if you can do it with a few million pounds of spending but can't do it with US$200m, then I think the jury is still out on the ability to buy a result.
But didn't the campaigns that spent significantly less money win in both Brexit and the US Presidential election?
1. If the services of Analytica/SCL are priced as per their price list so far and accounted for as a donation, Leave has spent more.
2. Quoting an old adage: "It does not matter how big your ****s is, it matters what you are doing with it". In this particular instance, the Trump campaign spent less because it spent it on "more advanced weaponry". Things will be back to normal the next election (unless there is a restriction of some sort on application of voter personal or aggregated data).
I know I am a couple of days late to this party, but cannon resist making an observation that some might find interesting.
The intent might have been to target those who otherwise did not bother to vote, but it is not clear from a comparison of the totals from the previous couple of presidential cycles that it succeeded for either of the two largest parties. The two-main-party vote was about 2 million more in 2016 than 2012, roughly in line with what one might expect given voting age population growth over the four years, and certainly not larger. It also was about a million fewer than in 2008.
On the other hand, over the two cycles, the smaller party vote grew from around a million (2008) to 2 million (2012) and near 8 million (2016), suggesting that much or most of the popular vote increase went to parties such as the Libertarians and Greens and was cast by those unhappy enough with both major parties to vote for a candidate with no conceivable path to an electoral win.
So money spent for such highly tailored adverts seems to have been largely wasted, especially since it is nearly certain that the Clinton campaign did it substantially more judging by the campaigns' total expenditures.
As has been revealed since (yes, I know I'm commenting on age-old "news"; just correcting the record), one of the main uses of this technology was to target indecisive voters ( = especially ones possibly leaning Hillary, I'd assume) and just throw so much confusing shit at them that they'd get frustrated with it all and not vote. An old technique, dunno if it was invented or just perfected by the Soviets, getting people to think there just isn't any truth to be had among all the contradictory messaging; perhaps that there is no such thing as "truth" any more.
So no: Fewer votes is not necessarily a sign that the CA-supported campaigning failed; on the contrary, perhaps more of a testament to its effectiveness.
Buried in the details - the constructed "news" postings that attract the interest of the supporters also have ads that attract their attention. Just viewing a webpage with ads can generate funds back, and if any of the other "news" articles in the sidebar generates more clicks, then the ads for those web views will also pay off for the posting corporation, the hosting corporation and the media in general. It becomes possible to generate funds to expand the work in progress, and to refine and elaborate on the work. These companies turn profits doing this work generating public support for Brexit, for the current POTUS, and for Western European white nationalism. That's my belief from reading about these corporations in different sources. If that makes them a good investment, they will also enrich players on the stock market, further cementing their positions in the status quo - they are becoming the "new normal news source" for millions of people on the Internet, who are also increasingly unlikely to trust or believe the major newspapers or television networks. The public stream of attention is being divided, without discussion, and then the new stream is being focused in support of certain policies and practices. It has now demonstrably worked at scale.
If you are still assuming there are only two parties, left vs right, then you are an idiot. There is only one party The Plutocracy, they like to dress up in two different outfits, each tailored to mindless morons, like yourself, who can't see beyond the media fairy-tale of the Repukelickants(tories) and the Democrabs(labour). They can spend huge amounts of money to make sure laws favor them and their corporations, who are now also people, thanks to morons like yourself who get nothing in return. Except perhaps the idea that your shitty, low-paying job is a "gift from your wealthy masters," or a tiny tax break that is a drop in the bucket for your wealthy masters. Face it, your grand plan to "fix the world" through Right-Wing, the tea party, and rich people "thinking" is as far-fetched as any children's story tale, and for an audience with a similar IQ and attention span. Far left, or far right, both are shitheads who can't manage reality. There is a shortage of smart people in the world who are beyond left/right, because morons can fuck and shit out babies more efficiently than normal people. You are living proof of this. In four years you won't be; rich, better off, or smarter. It is known.
>Is there any limit to the number of excuses the Left will try to come up with, to explain the fact that not everyone agrees with them?
What of are not of the left or right, but get royally pissed off with people spouting bullshit? How can democracy function in the way that we all agree it should, if people don't get called out for clear misinformation, be it in the side of a bus or in a Trump tweet? This is a tech blog, and physics and engineering doesn't give a damn what your opinion - or mine - is. When we have group problems to solve I hope we use empirical evidence to assess solutions.
In the past, wiser minds have created bodies such as the Office of National Statistics in an effort to prevent the subjective masquerading as objective. Without the ability to agree on what is true and what is not true, we're at the mercy of those who benefit from our division. Trump's tweets are so often demonstrably incorrect.
As for left and right, or in or out.... not everyone who voted had extreme views either way. In fact, a lot of people were pissed off with the shallowness of the Brexit argument from both sides. If I made a decision in good faith based in information that was later shown to be false, I would like an opportunity to fix the mistake. I would certainly be pissed off with some politician or tabloid expressing my weary ballot vote as being my sacred 'will' (as in 'the will of the people').
I wonder what their algos would make of Trump's Twitter posts. They seem more "candid".
You'd either get "input error, input garbled, please use a known langauge" or you'd hear the sounds of motors winding down, any lights on the computers would dim with many of them dropping out altogether as processor power is reduced, something akin to the first joke on this page.
(IIRC the version I read many moons ago spoke of gears winding down as the thing tried to reduce it's processor power, before asking "Did you watch the rugby last night?".)
... Theresa May's Brexit speech from January generates a 67 per cent openness rating, making her "liberal and artistic" rather than "conservative and traditional", and a 99 per cent score for her being a man. To be fair, a political speech is a controlled and stylised piece of writing...
To be fair, the software is based on evaluating truthful speech, and I reckon this just shows that Ms May is lying through her teeth, which is a fair assessment if you've ever heard her speak.
For the last hundred years or more, there has been more than a few people who're desperate to believe that magical mind control can be moved from myth and fiction onto some sort of scientific basis. See the book The Men Who Stare At Goats to see how deep the belief goes.
It's true that you can understand the people in ever greater detail with big data, but you only have to see the political upset on both sides when Brexit/Trump won to realise that there's no uber conspiracy here, just the normal fallible humans finding new and interesting ways to screw things up. Not that the changes being ushered in are necessarily bad for our deeply embedded political systems, but no-one could really claim that there is any evidence of a mastermind at work...
From: buzzfeed https://www.buzzfeed.com/kendalltaggart/the-truth-about-the-trump-data-team-that-people-are-freaking?utm_term=.pcNr7e0onN#.udr0bYr1o6
But interviews with 13 former employees, campaign staffers, and executives at other Republican consulting firms who have seen Cambridge Analytica’s work suggest that its psychological approach was not actually used by the Trump campaign and, furthermore, the company has never provided evidence that it even works. Rather than a sinister breakthrough in political technology, the Cambridge Analytica story appears to be part of the traditional contest among consultants on a winning political campaign to get their share of credit — and win future clients
back in the 1980s, I hit upon the idea that since UK elections are decided by a few thousand people, if those people could be identified, whoever had that data would have a lever into swinging General elections.
As I started, and mentioned it to my tutor, I had a visit from some very nice people, who suggested I research elsewhere. They asked that give them copies of my notes. They didn't ask that I give them the floppies which had the (Wordstar) originals, which hinted where they were from.
But ever since I used the phrase "abnormally large unrelated data sets", I have wondered where those notes got to.
if those people could be identified, whoever had that data would have a lever into swinging General elections.
Not any more. Our politicians no longer bother to put the effort into finding those people and persuading them to swing their way (if you see what I mean). Instead the politicians run after those people, asking them what they'd like their MP to believe, and dragging their parties with them. The consequence is that today's politicians are not leaders, they're followers. Nobody votes for followers.
It is another step toward algorithms making our choices...drip drip drip
There's some really fascinating discussion of this in Yuval Noah Harari's 'Homo Deus - a Brief History of the Future'. Trust me, read this and you won't feel at all comfortable by these sort of developments. It is truly one of the turning points in human history.
Algorithms are able to make much more accurate choices for you, from doctors diagnosis to choice of partner, if you want them to. Currently that is a choice. For how long, who knows. People are very willing to give up on making difficult choices. Take religion for example, for thousands of years people have 'followed the book', whatever one it might be. This is no different. People so often take the easy route because they don't want any make any effort themselves.
One thing that occurs to me is that ALL of these sort of algorithms that can have such a massive potential effect MUST be in the commons, open source if you like, and able to be influenced and updated my everyone with an interest. Unless that can happen we'll be at the behest of the exclusive owners of these algorithms, and you can be certain that their interests will be well apart from yours. Unfortunately I can't see any Gov't having the balls or the will to go down that route.
This post has been deleted by its author
While the technology used (and the psychological underpinnings) might not be the decisive factor that some - looking for easy answers - might latch onto, we shouldn't lose sight of its importance as one of the techniques used.
The Trump team (still waiting for more detail on the whole leave.eu thing) engaged in a staggering campaign of lies, lies and more lies. Being able to get the right lie, in front of the right target audience, at the right time, can't have been insignificant.
Where Cambridge Analytica come in, is in wielding one of the tools that they used to dupe the electorate and - yes - steal the election. (How can you put it any other way?) Outrageous bragging, threats, smears and deliberate lies gained Donald J Lincoln Kennedy Trump huge media coverage (a multiple of his rivals at each stage). Being able to repeat this rubbish to people predisposed to hear it had (and has) huge value to the 'new regime'.
However much importance pundits may attach to his populism, the elephant in the bathtub is the sheer scale, persistence and shamelessness of the lying, and how well it worked _and_continues_to_work_. Do you feel comfortable with this?
I think this has less to do with carefully targeting the right lies, and more to do with basic psychology.
In essence, people are more inclined to agree with things that feel, or sound right to them - this is a basic bias that is present in everyone (myself included). The problem is that reality is complicated, and things that are true don't often translate into simple soundbites in the same way that outrageous lies do.
The result of this is that printing big numbers on the sides of buses and implying that large amounts of money might go to the NHS gets people's gut feelings, despite being almost entirely devoid of any factual content.
The more subtle argument about how sending lots of money to the EU in 'membership dues' actually saves the country money that would otherwise have to be spent on things like paperwork, trade tariffs, coordinating international policing etc. etc. gets sidelined, and even worse, gets negated by the loud (and easy) shouting down and name-calling from those who are prepared to come up with those 'truthy' soundbites.
The cause of this is basic human nature - we have two decision making processes; gut feeling and critical thinking. One is easy (and often wrong), and the other is hard, and needs to be taught. Those who would like to control the masses also don't really like to have them doing too much of the latter, so critical thought is discouraged (historically in a religious context, and now increasingly in schools). Those who base their decision-making on facts and evidence are maligned as 'educated elites', 'liberal lefties' and 'so called experts'.
In this country we also have a culture of underachievement (those who try hard in school are bullied for it) so when we don't work hard, and expect a cushy job, we act all surprised when someone comes in from another country and undercuts us because they have a better work ethic. The 'easy' answer here is to blame immigrants, the fact-based one is that if we applied ourselves, we would be able to out-compete them, since we don't have the disadvantage of having to move to a foreign country where we don't speak the language. Obviously, this is an over-simplification; another element to this is exploitative employers who can more easily mistreat foreign workers who may not know, or be willing to apply their legal employment rights. Again, immigrants are not to blame here, and to suggest that someone from another country is coming here to 'steal' jobs is the sort of xenophobic drivel you'd read in the Daily Heil.
Anyway, I've wandered a little off-topic. My basic point here is that good decision making skills are not innate in human nature, we have evolved to make snap decisions and to be led by consensus. Whilst these traits can be manipulated, this can also be mitigated by teaching critical thinking to our children. If we did, we might not even have to point out that people should fact-check things they read on the internet...
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
