Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

People can be convinced they committed a crime that never happened (2015)

psychologicalscience.org

Post reply on HN
Print view View original

21–30 of 221 posts

Re: People can be convinced they committed a crime that never happened (2015)

#21

Post by nathan_compton »

DangitBobby wrote: post #12
nathan_compton wrote: post #5

Earlier quoted context omitted.

[flagged]

I just looked it up and it seems you are right. But I find it strange that "2%-10%" is considered rare, that's certainly more than I expected. 1. https://www.brown.edu/campus-life/health/services/promotion/...

From a Bayesian point of view, promulgating the false rape accusation idea is just irrational. When someone is raped (men and women are raped) they have a prior probability in mind about whether their accusation will be taken seriously. People they report to also have a prior as to whether the rape accusation is credible. In a world where most rapes are unreported there is almost no reason to adjust those priors towards "this is a false rape accusation."

Our legal system sucks and I think its bad that a person's life can be ruined or seriously damaged by a false accusation. But it is worse to be sexually assaulted and not have it taken seriously. I think if you balance it all out, focusing on the false accusation narrative is bad.

Re: People can be convinced they committed a crime that never happened (2015)

#23

Post by ekianjo »

> Evidence from some wrongful-conviction cases suggests that suspects can be questioned in ways that lead them to falsely believe in and confess to committing crimes they didn’t actually commit.

It's been known for like dozens of years at least. What happened in soviet Russia and in China during the cultural revolution was all about making people convinced that they had committed crimes and that they deserved their punishment, and it worked very well.

Re: People can be convinced they committed a crime that never happened (2015)

#24

Post by nonameiguess »

This was published 8 years ago. Makes me wonder if it's replicated yet or if anyone has tried.

I certainly am no fan of the reliability of human memory. It surprises me that more people aren't skeptical just from their own experiences. So many times I've told stories of things I did or remember from when I was young, especially to my wife, only to eventually visit my parents and find out I'm wrong and that never happened.

And, of course, among my earlier memories are from kindergarten. I remember seeing the Challenger explosion happen in class and I remember missing half the year because of pneumonia. They can't possibly both be true because they'd have happened at the same time, but I remember both nonetheless. Is my memory of watching the OJ verdict live in high school history class also false? I have no idea, but I wouldn't be the least bit surprised. I have no idea how much of my remembered life actually happened without independent corroboration.

It leads to such conflicted feelings when the subject comes up of people remembering childhood abuse they'd "suppressed." No doubt many more children are abused than remember and many, if not most, abusers never come to justice. But also no doubt many of these memories, like the entirety of the Satanic Panic, never happened, and many people are in prison or registered for life unable to get a regular job, who never did anything.

It's worth remembering this as we see the development in real time of LLMs and other somewhat intelligent tools that are able to just make shit up and seem totally confident about it. They're not the only ones that hallucinate. All prediction and recall systems that are inherently probabilistic but need to tell only one story are likely to be susceptible to this kind of thing, including any and every AI we will ever develop, including ourselves. Claims need to have evidence and be independently verifiable or you shouldn't trust them. Even your own claims.

Re: People can be convinced they committed a crime that never happened (2015)

#25

Post by localplume »

madaxe_again wrote: post #14
joelegner wrote: post #3

Simone wielding AI nefariously will eventually use this tidbit about human nature to terrible effect.

AI is not necessary to make your population believe the Big Lie - in fact, by applying a consistent false narrative to an entire group, you get reinforcement effects by more impressionable members of the group towards those with a greater resistance. This has been used, repeatedly, through history. Now, where AI could be applied would be in having a tailored, sculpted false narrative for each and every subject, which…

You touched on a disturbing point. People like to imagine that some govt will wield to indoctrinate the populace, but it is more likely this kind of tool would be used for marketing or advertisement. look at how many people gladly work for FANG/whatever companies and develop tools just to supercharge advertisement but see no moral issues with it. as long as they're making money, and a lot of money, who cares right?

I mean advertisements already are basically brainwashing right? quirky jingles or catch phrases associated with some food or product, that stick in your head.

Re: People can be convinced they committed a crime that never happened (2015)

#26

Post by Aerroon »

I wonder if this is why people can become angry and aggressive when you accuse them of something. That would provide a natural defense against this kind of suggestibility.

It obviously wouldn't work with the modern police, because they would overpower you, but a few thousand years ago? I think it could've worked.

Re: People can be convinced they committed a crime that never happened (2015)

#27

Post by fzeroracer »

nathan_compton wrote: post #21
DangitBobby wrote: post #12

Earlier quoted context omitted.

I just looked it up and it seems you are right. But I find it strange that "2%-10%" is considered rare, that's certainly more than I expected. 1. https://www.brown.edu/campus-life/health/services/promotion/...

From a Bayesian point of view, promulgating the false rape accusation idea is just irrational. When someone is raped (men and women are raped) they have a prior probability in mind about whether their accusation will be taken seriously. People they report to also have a prior as to whether the rape accusation is credible. In a world where most rapes are unreported there is almost no reason to adjust those priors towa…

I don't think this argument is worth continuing. Not in the sense of being wrong; your argument is correct and I agree with you. But in the sense that the site that started this whole thing is insane. If you take one glance at the rest of the site it becomes quickly clear that the person managing that site is not well.

Re: People can be convinced they committed a crime that never happened (2015)

#28

Post by yosito »

> The fact that the students appeared to internalize the false events to the extent that they did highlights the fundamental malleability of memory

This seems like a bit of a leap to assume that this is about malleability rather than just lying to researchers and saying what they feel is expected. I could very easily imagine myself playing along with someone trying to convince me of false memories just because I don't want to put up a fight against someone who is gaslighting me.

Speaking of which, how is gaslighting students considered ethical research?

Re: People can be convinced they committed a crime that never happened (2015)

#29

Post by im3w1l »

localplume wrote: post #25
madaxe_again wrote: post #14

Earlier quoted context omitted.

AI is not necessary to make your population believe the Big Lie - in fact, by applying a consistent false narrative to an entire group, you get reinforcement effects by more impressionable members of the group towards those with a greater resistance. This has been used, repeatedly, through history. Now, where AI could be applied would be in having a tailored, sculpted false narrative for each and every subject, which…

You touched on a disturbing point. People like to imagine that some govt will wield to indoctrinate the populace, but it is more likely this kind of tool would be used for marketing or advertisement. look at how many people gladly work for FANG/whatever companies and develop tools just to supercharge advertisement but see no moral issues with it. as long as they're making money, and a lot of money, who cares right? I…

If you google Paris and you get an ad for a hotel, that isn't brainwashing. It's fun to be radical but let's stay grounded here.

Re: People can be convinced they committed a crime that never happened (2015)

#30

Post by cplusplusfellow »

Aerroon wrote: post #26

I wonder if this is why people can become angry and aggressive when you accuse them of something. That would provide a natural defense against this kind of suggestibility. It obviously wouldn't work with the modern police, because they would overpower you, but a few thousand years ago? I think it could've worked.

I can hear the police response now. “Why are you so angry about me accusing you of committing such a heinous crime that may even bear the death penalty if you didn’t do it?”
Post reply on HN