[numpad0]

Being lawful-evil person I am, I once imagined a “news” app such that display content or suggestion ever so unnoticeably slightly change depending on how user behaves judged by device camera.

A facial expression ML model would be trained beforehand so camera data would not have to be collected, but only pushed to users along app updates, and content would be analyzed and encoded at server side to determine what users are to feel, along with hints to aid learning and correlation between different stories, all subconsciously and automatically.

As for what to do with a system like that...I don’t know, make boatloads of cash by forcing in-app purchases? Push your favorite but little known novels and comics? Make group of people I like by my specifications? Promote good posture, regular exercises and force people to look up? I wouldn’t want to use it so specific elected officials pass or resign, or world’s atrocities to become accepted...

Then I got bored. But could it have been something like this?

[SkyMarshal]

I wonder if people make distinct enough facial expressions while reading online content for this to work. For myself, I pretty much stare blankly at the screen the whole time without much change in my expression. The ML would have to be very good at detecting slight nuance.

[hairofadog]

To your point, one of the things I worry about with apple news, for example, is that the underlying ai won’t understand _why_ I dislike a particular news story. Generally I downvote hyperbole, clickbait, and paparazzi content, but I would guess it’s more common for people to downvote things based on the substance of a story, e.g. downvoting a well-written, truthful article because the truth in that case is inconvenient for their political stance, or because it’s about a loathsome-yet-newsworthy person. A facial recognition system would have to know the difference between my face being frustrated at a high-quality story about the fires in the western states vs my frustration that apple news seems to find Andy Borowitz to be hilarious and relevant. Maybe possible, but lots of nuance under the surface.

[AndrewThrowaway]

While your face might seem still, e.g. your heartbeat, your eye movements can reveal a lot.

I remember how I was blown away by this video a few years ago:

https://youtu.be/3rWycBEHn3s?t=171

[thrav]

Wasn’t there a proof of concept for exactly this about how social media apps were probably already doing this? I believe I saw it on HackerNews. It was a site where you gave them camera access and then ran something like a 60 second experiment. Seemed academic.

[almyndz]

yeah, usually when I'm reading the news I'm only screaming on the inside

[heartbeats]

The real question is why nobody bothers. Why doesn't Facebook psychologically manipulate their users into being OK with tracking?

[rmrfstar]

> Being lawful-evil person

The word you are looking for is "sociopath". Oliver Wendell Holmes' "bad man" is an analytical tool, not a recommendation.

There's nothing inherently evil about the mood-detection feature you have described, provided it is deployed with continuing and informed consent from the user.

The feature is already deployed on billboards in public spaces, although this instantiation is evil as "users" are generally unaware of it, have not consented to participation, and cannot opt out.

[Raphmedia]

Please don't throw mental health diagnosis willy-nilly around on the internet. Not only is it impossible for you to make such a diagnosis, it is rude to both the community you are using as a punch line and to the poster.

[numpad0]

Existing implementations only detect moods and collect data for analysis, my idea is if you're just going to do open-loop brainwashing, you don't have to know users at all so it could be done without "invading privacy".

> sociopath

ok, I mean no harm to humanity, but, um,

[rmrfstar]

I was making a general statement that the ideology of "maximize profits by any lawful means" is a profoundly antisocial one.

I don't think many people actually subscribe to that ideology, but enough do that it's worth challenging whenever it is articulated (as it seems to have been in your comment).

[rocketmaster1]

> ok, I mean no harm to humanity, but, um, Sorry to gate keep but then you're not lawful evil. A lawful evil character is completely okay with harming people (through legal means of course) as long as it benefits them.