Live data from Hacker News

Should There Be Limits on Persuasive Technologies?

schneier.com

21–30 of 241 posts

Re: Should There Be Limits on Persuasive Technologies?

#21

Highly interesting topic, I am currently planning to write my thesis about persuasive technology. If anyone has articles/ ideas/ stories/ books or papers to share I would highly appreciate it! I thought about researching how humans react to persuasive technology and start self regulating by installing adblocker/deleting apps... I am all ears if there are other interesting questions you might think of :)

I should think that actually having academics access to the internal effectiveness metrics would in and of itself be hugely beneficial.

- what is the range of ads to be served like? I rmeber thebfounder of duck duck go saying lack of PII made little difference to him - someone is searching for "PC Monitor" then the range of ads seems pretty obvious. Is this true in "persuasive" mode? Are there bidders trying to serve political ads or lingerie when I search for a monitor? How can you tell?

- is this a zero sum arms race ? A study I cannot find should that Political races are decided by who the public prefers not by total spend. Congressional races showed that if ad spend was matched the winner stayed consistent - it's only when one side out spent the other did ads have an effect. If I am outraged by ad X do ai also get served the anti-X advert? Would it work if I was? (this is the equalicalent of how many adverts saying fruit and vegetables are good for you outweighs adverts for sugary drinks? There is probably a lot of metrics out there on that - how comparable is it?)

good luck

Re: Should There Be Limits on Persuasive Technologies?

#22

I keep coming back to my default idea here - that PII needs to be seen as legally owned by me and only licensed to others for use. The default legal framework should include medical / epidemiology research as freely licensed and commercial use as ... well let's just say i think my license conditions will be expensive. If an advertising channel then shows ads that breach license they are liable. A fairly simple licens…

If you're a US citizen with a Facebook account, your value for Facebook is about $ 200 per year. That's the average, including children, seniors, etc. If you're a tech worker in your 30s, it's probably 3x as much. If you want to use Facebook without targeted advertising, you need to either convince them that they don't need all that money, or pay it yourself. And that's just Facebook. In other words: the internet eco…

Worth noting that the $200 ultimately come from our pockets. We are paying those $200 anyway through markups on products ads made us buy.

Re: Should There Be Limits on Persuasive Technologies?

#23

Earlier quoted context omitted.

If you're a US citizen with a Facebook account, your value for Facebook is about $ 200 per year. That's the average, including children, seniors, etc. If you're a tech worker in your 30s, it's probably 3x as much. If you want to use Facebook without targeted advertising, you need to either convince them that they don't need all that money, or pay it yourself. And that's just Facebook. In other words: the internet eco…

Worth noting that the $200 ultimately come from our pockets. We are paying those $200 anyway through markups on products ads made us buy.

Isn't the 200 bucks more a way for FB sellers to get your money instead of other shops?

Re: Should There Be Limits on Persuasive Technologies?

#24

Earlier quoted context omitted.

If you're a US citizen with a Facebook account, your value for Facebook is about $ 200 per year. That's the average, including children, seniors, etc. If you're a tech worker in your 30s, it's probably 3x as much. If you want to use Facebook without targeted advertising, you need to either convince them that they don't need all that money, or pay it yourself. And that's just Facebook. In other words: the internet eco…

Worth noting that the $200 ultimately come from our pockets. We are paying those $200 anyway through markups on products ads made us buy.

I would love to see a correlation score between profiles and purchases

Re: Should There Be Limits on Persuasive Technologies?

#25

Earlier quoted context omitted.

Worth noting that the $200 ultimately come from our pockets. We are paying those $200 anyway through markups on products ads made us buy.

Isn't the 200 bucks more a way for FB sellers to get your money instead of other shops?

No, because it's $200 on top of what you would otherwise spend on the stuff you buy. The cost of marketing has to get baked into the price of goods or the sellers would be losing money. If they weren't advertising, or could advertise for less, they'd be able to lower prices.

Re: Should There Be Limits on Persuasive Technologies?

#26
I wrote this for a related article yesterday:

---

If we know that humans have all sorts of cognitive biases, how come it's ok to use that fact while at the same time we insist there's some kind of free market?

Say you discover that putting good-looking women next to cars causes the sale of cars to increase. Why does nobody question whether it is legitimate to do so? It's as if there's a line between actively lying ("Studies show that men who buy this car will find many many women attracted to them") and just putting it there suggestively, for some as yet undescribed but working cognitive bias to do its magic.

Some advertisers even make a joke out of it, eg the Lynx ads where the dude is thronged by a huge horde of women. It's a cliché, for a good reason.

I suppose most people will just say you have free will and it's your own fault for thinking what was suggested, but I sense this is more of a grey zone than most people are willing to admit. How can the free market work if everyone is so easily affected by suggestion?

---

Of course this also applies to the free market in ideas. In what sense are people free to make up their minds if it's decided for them what they should see, whether or not the government is doing it or FB? Isn't this the same as the authoritarian nightmares that we've been pointing fingers at?

Re: Should There Be Limits on Persuasive Technologies?

#27
> The technology exists to take your likeness and morph it with a face that is demographically similar to you. The result is a face that looks like you, but that you don’t recognize. If that turns out to be more persuasive than coarse demographic targeting, is that okay?

I have wondered if in the future, movies/TV shows will be personalized with your name. I have found that when I watch a movie where the main character shares my name (and therefore other characters say the name a lot when talking to/about him), it makes the experience more immersive. It would seem fairly trivial to substitute in other names in a pretty smooth way (it would get tougher if you had to also adapt things like a business card that is visible on-screen, however).

Of course, this could be taken to the next level if you changed the faces of the actor via deepfake-like technology. I don't know how actors would feel about this sort of thing, but hey maybe it would open up the door for a bunch of new actors, who would essentially be a blank slate for customized faces. Imagine a world where Hollywood actors don't have to be good-looking!

Re: Should There Be Limits on Persuasive Technologies?

#28

Earlier quoted context omitted.

If you're a US citizen with a Facebook account, your value for Facebook is about $ 200 per year. That's the average, including children, seniors, etc. If you're a tech worker in your 30s, it's probably 3x as much. If you want to use Facebook without targeted advertising, you need to either convince them that they don't need all that money, or pay it yourself. And that's just Facebook. In other words: the internet eco…

Worth noting that the $200 ultimately come from our pockets. We are paying those $200 anyway through markups on products ads made us buy.

Not necessarily. I don't know what percentage of advertising revenue is from political advertising, but that's money that Facebook makes off of you that doesn't come out of your pocket. Essentially any time you're shown an ad that aims to convince you to support some idea, or vote for something, or change your behavior in some way (like going vegan) instead of trying to get you go spend money, you provide value to Facebook that doesn't come out of your pocket.

I suspect that this is a lot of the value of "the next billion". Why bring Facebook to people who have little disposable income? Because there are still ad buyers who will pay to reach those audiences if they have votes to give, or even just public support for some cause or another.

Re: Should There Be Limits on Persuasive Technologies?

#29

> The technology exists to take your likeness and morph it with a face that is demographically similar to you. The result is a face that looks like you, but that you don’t recognize. If that turns out to be more persuasive than coarse demographic targeting, is that okay? I have wondered if in the future, movies/TV shows will be personalized with your name. I have found that when I watch a movie where the main charact…

I can already imagine movies and tv shows starting, Please tell us your name out loud and then some deep fake voice magic will adapt the actors voice so it pronounce your name the way you pronounce your own name.

That’s has to be so weird to hear your name in shows/movies.

Re: Should There Be Limits on Persuasive Technologies?

#30

Earlier quoted context omitted.

If you're a US citizen with a Facebook account, your value for Facebook is about $ 200 per year. That's the average, including children, seniors, etc. If you're a tech worker in your 30s, it's probably 3x as much. If you want to use Facebook without targeted advertising, you need to either convince them that they don't need all that money, or pay it yourself. And that's just Facebook. In other words: the internet eco…

Worth noting that the $200 ultimately come from our pockets. We are paying those $200 anyway through markups on products ads made us buy.

Potentially buy.
Post reply on HN