As a neuroscientist, I can say that calling it "neurotargeting" vastly, vastly overstates the contribution of "neuro" to these ads. Advertising has had more than a century of practical experience in manipulating the human mind. fMRI studies have added approximately zero to this understanding.
Really, fMRI studies are their own form of neurotargeting, though aimed at ad buyers rather than ad consumers. Pretty colorful pictures of the brain convey the idea "Gosh, we're so smart and advanced and differentiated from those other ad agencies", even if those pretty pictures provided basically no new insights into how to sell anything.
The types of "neurotargeting" describes in the article - using data from emojis, personality tests, OCEAN models, fMRI studies - are all very noisy compared to what actually does work in advertising. You perturb the ad shown, measure conversions, and then just directly optimize for conversions. It's a variant of A/B testing described in the article, but you aren't comparing just two variants and having a human decide based on metrics, you're comparing thousands of variants and then training a machine learning model to predict conversion events.
All of the major online ad networks - Google, Facebook, etc - actually function like this, and have for the past 15 years. And it is effective to the tune of about $300B/year, looking at Google's ad revenue.
Barely related, but do you have advice for someone who wants to work on a very specific project at Google? I'm not at the point of hireability yet, but in the long-term I want to optimize my chances to get hired to work on AndroidXR. Should I just put my effort into writing tangentially-related open source projects? Should I be trying to grab positions at similarly-focused companies first? I don't really care about optimizing for comp, and I don't even really mind if I'm only contracting and not at Google proper, I just very specifically want to work on that project.
It would be > 0, but pretty significantly < $300B.
I think a lot of people underestimate the importance of good targeting to the online ads market. It's taken over from print/media/TV brand advertising because it's orders of magnitude more efficient. When Google was young (~2005) advertisers were reporting that they'd get 3-5x more conversions from Google at 1/10th the ad budget because the conversion rate was that much higher.
I think it's "neuro" here is more "psychological", meaning the ability to automatically know en masse the people. In the past we have had ads almost "for everyone", it was not much possible choose a specific cohort with a specific message, yes you can send some brochures in specific places normally frequented by a certain cohort of people, you can use friends, but it's still not much targeted.
Nowadays we are able to "know automatically" and separate specific cohort of people, let's say we want something pro-rail and we have some angry users waiting for a late train, a news arrive in their aggregator narrating another train delay somewhere else caused by a landslide, another one caused by a suicide, finally an article on the terrible current years of rails due to climate change and social unrest. The angry target probably remain a bit angry, but far less than before. Similarly in a campaign against rail another news can pops up just when someone look for a possible trip/to buy a ticket, about some horror story of passengers leaving at 40+℃ for some hours due to some malfunction and bad management plus another about an accident due to little maintenance and safety checks.
Nothing new under the Sun, but the ability to get the right news/ads at the right time makes the level of effectiveness something much, much bigger than ever and the ability to target users is still in very few hands. They can't "rule the society" but being able to steer they are kind of ruler anyway.
The term "neurotargeting" is the kind of manipulative framing they notionally abhor, there is nothing "neuro" about it. This is just classic computational propaganda but I guess that term isn't edgy enough.
They greatly underestimate the difficulty of the computational problems surrounding automated production of effective and highly granular propaganda at scale. Targeting is crude and coarse because that is a technical limit of off-the-shelf analytical processing platforms.
The free will debate usually falls into two camps. One holds that we have free will, period. The other holds that it’s some kind of illusion, period.
I believe that we have it but it is not some kind of context free limitless divine power. It’s constrained like all other real things.
Our free will is more like the ailerons on a plane. It can steer the plane. It cannot make the plane stop instantly or make 90 degree hard turns. It cannot stop another plane from running into you, or stop a bird strike, or rescue the plane if the wing falls off.
I think what I’m getting at is that as our technology for influencing each other advances we can never be sure exactly where we are in comparison to the strength of our own free will. At what point might we invent ways to influence each other that renders us mindless biological robots?
That’s an interesting way to put a floor on it but I think he’s neglecting emergence and nonlinear dynamics. Deterministic objects in a system can become computationally irreducible and manifest higher order behaviors that are not reducible or predictable.
Of course then you get into the challenge of defining free will. Is it enough to be impossible to predict or is actual acausality required? If the latter you are into quantum mechanics or bust.
> they understand, from centuries of experience, that provoking strong emotional responses is one of the most reliable ways to get people to change their behavior
this is easily defeated by teaching everyone a sprinkle of meditation (it has been quite shocking to me the extent to which I'd have an emotional reaction -> rationalized argument -> action, even over things like, slack messages from coworkers/my manager. I used to think it was always rational thought -> action. But noticing this makes it very easy to alter the pattern)
now I want to see a response article: "Meditation will save democracy"
Ironically/paradoxically, a graceless or haughty (over)correction can undermine our own efforts to correct this in ourselves and especially in others, so I would agree that meditation and general mindfulness practice as well as empathy are helpful in understanding and empowering us to overcome these hurdles in interpersonal dynamics.
Jan 6 is unlikely to have been unrelated to “data-fueled neurotargeting” (title is over dramatic imo).
First impression information really sticks in people’s minds (I heard x and then you tell me y in conflict with x, I will believe x more easily). And targeting is doing this daily.
Republican voters talk more about the democrats than democrats do, democratic voters talk about republicans than democrats do. Obsession with the other is what happens in the current gridlock, and the most unproductive congress in history.
My theory on Trump supporters is that it's confirmation bias and sunk cost fallacy.
Centrist Republicans and independents supported Donald Trump in 2016 because Hillary Clinton was a weak candidate. They may not have agreed or liked Trump as a person, but they voted for him.
But then, as Trump said and did increasingly extreme things, they had to create a narrative that post-hoc justified their support.
And he made 'The world is out to get him. He's been treated unfairly. Everyone else is lying' easier than 'I was taken in by a con man, who turned out to be even worse than I thought.'
Which is how 'I don't like the man, but I support his policies' voters became 'Everything about Jan 6 was fine' voters.
And why I try to inculcate a 'I am not a member of any political team or identity' in myself. Politics should be about policy, not brand.
You require a majority to institute a policy. Practically zero people believe in policies identical to yours. The only way for you to get any of your policy preferences is to find people whose policies you can live with, and who can live with yours. You will usually disagree with them, but they're better than the people who hate most of your policies and will do exactly the opposite.
That fact would seem to make democracy barely tenable at best. It becomes entirely untenable when you can't even identify your own brand any more.
There are tens of millions of people who would say "Of course I would never vote for a felon who attempted to overthrow an election." But they will find a "but...", for reasons that are objectively irrational but completely coherent in their own heads.
I point that out because, whatever policy you want, you're going to need brand-mates. A lot of them. The more you need, the more you people you'll dislike having on your side. But it's unlikely to be beneficial to say, "Forget it, I'll take the greater of two evils."
Flags won’t hoist or wave themselves, red flags doubly so, and they’re a useful determinant of which way the political winds are blowing, and whether they’re shifting.
Another movie from that same era I enjoy is The Pentagon Wars about the development of the Bradley Fighting Vehicle. It’s one of the better made for TV movies I can remember.
complicating matters is that the world IS out to get him. it doesn't actually matter all that much that he deserves it, especially since to his followers, pretty much everybody in power deserves it, so the fact that all this stuff is only happening to their guy proves to them that the thing you ACTUALLY go down for, is being their guy.
further, if you trust the glowies post-Snowden, idk what to tell you. Trump said he'd drain the swamp, and the swamp fought back and won.
unfortunately for people who do believe in the corruption of the three letter agencies, Trump's tactic of undermining democracy itself (however much we do have) in order to appease his narcissism is actually worse than the career spooks in control of the government (and much of the media narrative; the FBI Agent / CIA analyst to "political contributor" pipeline is not hidden at all, you only need pay attention to the people the major news networks call "experts" on all sorts of topics)
It’s shorthand for agents of the state and those who do their bidding or act on their behalf directly or by proxy, especially federal agents. It was coined by the TempleOS guy, apparently.
> The term was coined by computer programmer Terry A. Davis, who allegedly believed that the CIA was stalking and harassing him. "Glowie" is often used in online forums to refer to government agents, especially undercover operatives who infiltrate online extremist spaces.
> "Glow in the dark" and its derivative terms have been used to refer to various groups: newcomers that do not fit in with the culture of certain forums and are thus suspected to have bad intentions, journalists who report on extremist groups, tech companies that collect users' personal data, and others.
Jan 6th was done via what as near as I can tell are cult indoctrination and mass scale rabble rousing techniques as old as civilization. If newer tech like this was used it was to boost the effect.
I think it’s possible that the I AM cult was the template for some of it like Qanon. Look into the tail end of that cults saga when they tried to set up a march on DC to set up their cult leader as king. Flynn and a few others have made direct references to it such as Flynn’s weird I AM inspired prayers at Q events.
As Hubbard said… want to get rich and powerful, start a religion.
us vs them has an ad cross section too! … read the article?
Think about the last time you used someone else’s device in your family, or watch them use it… listen to the “news” they get and ask if they get different news than you. Ask why they believe what they believe, unjudgementally and with curiosity. This fact finding mission will surprise you.
It might be true certain people are less discerning than you, but I can guarantee that the small difference in discerning is not a linear effect in the distortion they see the world with.
I see bullshit across the political spectrum, but the Q cult was kind of unique. I don’t see an obvious left wing analog at least today in the US.
If you consider totalitarian communism left wing then you can see analogs in Maoism and Leninism. That was a long time ago in other countries though. The US seems to have a more vibrant totalitarian cult scene on the far right today than on the left.
This is what Harari and others have been warning for a while. Once you build profiles for everyone, you can craft stimuli that controls their behavior and in that way rule them.
Maybe we just stop viewing the internet as the ultimate human end and return to the culture that existed for millennia before: local interactions; live performances; physical media; in-person elections on a single day with physical ballots.
I think the authors overlook a more serious turn of politics in the past 50 years: politicians have monotonically been hoarding powers in the name of safety , technocracy or simply "trust us bro". This leaves the voting population unaccountable for their own choices, since none of their bad choices was allowed to affect them negatively. This made people vote increasingly with Pathos, aware that there will be no real consequences, or they will become a burden for the next generation.
> people vote increasingly with Pathos, aware that there will be no real consequences
Isn’t this the same as it ever was? In this view, voters deserve what they get, yet are unlikely to get what they deserve, but should we will it different? Ought not a tyranny of the majority set the stage a for tragedy of the commons?
I find myself noticing some confusion while reading this article (and several previous articles along the same line).
This seems like "neurotargeting" could be a danger to a very specific conception of democracy, but not a danger to democracy in the way that people, in my experience, generally use the word. And certainly not a danger to democracy in the way that people such as Edward Bernays would have thought about it.
For complete clarity, the way I've generally heard people use the word "democracy" is (roughly) regarding a system of governance where the will of the people is used to decide legitimacy, either by voting directly on laws or by voting on representatives to govern for them.
It's not automatically clear that changing the will of the people is actually a danger to that system or makes it any less legitimate (assuming the will of the people itself is what you value). And it seems weird that this would be a danger.
Is this just a different definition of democracy (possibly a regional variation?), or is it actually being used in a non-standard way here, because it seems like this would be a much bigger threat to liberalism than democracy (is democracy a short-hand for liberalism in some circles?).
A critical property of active neurodefense is denial of emotion sensing to the adversary. This means no iris sub-second response data, facial expressions, or heart rate via cameras and wifi radar sensing via local AI/ML inference. Use an OS which keeps sensor data on-device, via some kind of security enclave.
> To better understand neurotargeting and its ongoing threats to democracy, we spoke with one of the foremost experts on the subject: Emma Briant, a journalism professor at Monash University and a leading scholar of propaganda studies.
Actually, this sounds instead of the best way to understand the subject, the best way to get a very narrow and biased view and fear mongering view. Basically, this is someone whose whole career is dependent on convincing people that there is a problem whether or not there is.
It's been nearly 8 years now and the elite class still seems reluctant to come to grips with the fact that they lost an election to a boorish gameshow host not because of foreign adversaries or clever marketing but because many people not in their social circle decided they preferred him to another political dynasty that didn't appear to represent their interests -- and in fact made it a point to openly despise these voters publicly by calling them deplorable. It isn't rocket science.
Granted, Trump made it easy to point the finger at him, but agreed that the lack of Democrat self reflection over how the uglier impacts of globalism on the middle class drove the 2016 election result was ridiculous.
And similarly, how the taken-for-granted poor rural and evangelical vote upended the Republican party in the primaries.
Political parties ignore blocks of constituents at their peril -- and that's always been the strength of democracy.
As I recall, she was actually pretty specific in who she called deplorable, and it was an apt description. "They're racist, sexist, homophobic, xenophobic, Islamophobic". Saying that was half of Trump supporters was hyperbolic and perhaps inaccurate, but MAGA world adopted 'deplorable' as a badge of honor of their own free will.
Do you think they indeed self-identify as racist, sexist, homophobic, or Islamaphobic? I can only speak for the people I know personally, but that label does seem to fit. When you think you are being oppressed by "the alphabet mafia" you may want to take a step back.
>and in fact made it a point to openly despise these voters publicly by calling them deplorable.
Nope. Hillary Clinton was specifically calling out the racists and bigots within the Trump camp, not all Trump supporters. Which is obvious to anyone who bothers to read the actual quote. It isn't complex or nuanced, or even that long.
And I don't know why Trump supporters still insist on believing otherwise, since one would expect that they would want to distance themselves from the bad elements within their party. I can only assume they find solidarity with such elements to be more important than conceding that a Democrat might actually have a point.
Ah well. You sleep with the pigs, you wake up smelling like shit.
She did say half, which may be an overestimate, but she then provided a description. The rest of the supporters gleefully adopting the label, knowing the definition, is ... interesting.
It's hyperbole, and not mathematically accurate, but not entirely wrong. That white supremacists flocked to Trumpism like a moth to a flame is a well documented phenomenon, and the preference of Trump supporters to prefer to blame the messenger is telling.
>Enfranchisement for women and racial minorities in the United States took centuries to codify, and these citizens are still disproportionately excluded from participating in elections
No acknowledgement whatsoever of poor people. Clearly this is a trash article.
Democracy is already dead. As in who spends more or buys more Ads wins elections. For decades now.
Trump and the Brexit vote broke the traditional media-advertising complex but that got fixed fast.
American democracy is already all but dead. The reason being that lies fuel the decision making process: from campaigning to lobbying.
They claim it’s democracy when elected officials vote on bills that are largely against the interests and desires of even the people that elected those voting representatives.
I think we can recover though.
What we need to do is extend the powers of our justice by jury system.
Bills in congress should be voted on by a panel of jurors conducted by a judge.
The elected officials are there to litigate either for or against, but it is the jury’s ultimate decision.
I rarely think I have anything to offer the word of civics as it seems to me to be motivated in ways antithetical to my own desires in governance, but I truly think this is an idea that will have a very positive effect on the current failures of the American system of democracy.
My personal twist, if I were building a system in a first-world country, would be to have a bicameral legislature.
Upper house has 4 year terms with term limits (say, 20 years max) and has professional politicians.
Lower house is elected for 4 year terms by lottery, a portion turning over every year, in the same way as jury duty. Then support them with a strong research civil service (e.g. the CBO).
Juries are terrible at offering specialized knowledge, which is becoming increasingly critical in the world.
That's why their primary use in the US (trials) offloads all the specialty requirements onto other parties (adversarial prosecution and defense teams, judges, court staff).
Consequently, in using juries you gain some measure of protection from corruption and responsiveness.
Congress members are rarely experts (but those that are do exist), but their staff usually includes some.
And the problem with expert witness testimony is that it needs to be validated.
If you know nothing about a topic, and I and someone with an opposite viewpoint both put experts in from of you that support our positions, how do you decide?
Better funding and expanding a non-partisan, Library of Congress type resource seems like a better approach to me.
> If you know nothing about a topic, and I and someone with an opposite viewpoint both put experts in from of you that support our positions, how do you decide?
How indeed. How, for instance, would a congress member? What about staffers make them better experts than expert witnesses in a trial? Who are these expert staffers accountable to?
I feel you are being a little unbalanced in your criticism.
I'm all for better "resources", but fail to see how that could be held up as an "either or" replacement for my proposal.
Also, I am confused how it would remedy the ills I am addressing with the proposal. There's a meme online to the effect ~"remember when before the internet we used to think hate was result of lack of access to information".
Your representative will still have their staff and likely even their token experts, but it will still be on them to plead the case to a jury.
I feel those of us with interest in conjuring math in stone with electricity tend toward thinking there is the binary of right and wrong. That, "if only we had more data everyone would agree", but humans and human societies are much more complex and often an argument can be made for both sides of an debate regardless of how sure I am of my side of it.
The point is how do you ensure the argument with better evidence is the one we use to determine policy.
While only using existing tools from our current system I would argue the best way we have is trial by jury.
how are the jurors chosen? I'd expect this would be too small a sampling of people to be really representative, and I think it's often easy to confuse jurors who are (by definition) not experts.
Not that I think they system as it is is great, but I'm not sure this helps / doesn't have its own significant failure modes.
Really, fMRI studies are their own form of neurotargeting, though aimed at ad buyers rather than ad consumers. Pretty colorful pictures of the brain convey the idea "Gosh, we're so smart and advanced and differentiated from those other ad agencies", even if those pretty pictures provided basically no new insights into how to sell anything.