Dopamine: An AI platform for designing human behavior • Hacker News Board

Re: Dopamine: An AI platform for designing human behavior

#141

Post by 013a » Thu, Nov 09, 2017, 6:20 PM UTC

TimTheTinker wrote: post #19

013a wrote: post #9

The same company sells an api to help make your applications as addicting as possible [1] and a service for users to help control their addiction to applications [2]. This organization is disgusting and is evidence enough that our industry has no sense of ethical responsibility. When massive regulation lands on Silicon Valley and we whine about the impact it has on innovation, remember companies like Dopamine Labs wh…

To be fair, all the case studies they list [1] utilized their service to improve users' personal habits -- diet, exercise, etc. Behavior shaping isn't necessarily morally wrong to use, though companies like Facebook and Google are almost completely incentivized to use them against users. [1] http://www.usedopamine.com/assets/pdf/Dopamine%20Labs%20Case...

Intentional behavior shaping is a light form of mind control, and its very much a moral gray area. You can totally use mind control for positive things; that doesn't make it alright.

Re: Dopamine: An AI platform for designing human behavior

#142

Post by 013a » Thu, Nov 09, 2017, 6:36 PM UTC

mmayberry wrote: post #135

IanCal wrote: post #100

Earlier quoted context omitted.

Why not people wanting to shape their own behaviour? Nudges to exercise more from an exercise app I've installed is hardly diabolical.

you are correct. there is nothing wrong with giving humans the tools they want/need to be the better versions of themselves.

This is such a fine gray line that its impossible to conclusively categorize something as evil vs good.

For example: If I decide I want to exercise more, maybe I would appreciate an app which helps me become addicted to exercise. That might be good for me.

However, what if the app pushes me to exercise too much, and I begin to experience health problems associated with destroying muscles? Or, what if the app is formed on bad exercise science and it suggests routines that are bad for me? Now suddenly the addictions created by the app are working against me; the app isn't being irresponsible by helping me exercise, but it is being irresponsible by modifying my behavior and decision making processes to favor using it.

Similarly, is Instagram "good" for you? Probably not. But, per your comment: "there's nothing wrong with giving humans the tools they want". People might want to be addicted to Instagram; that doesn't mean it is good for them and that Instagram should deliver. People want to be addicted to nicotine and alcohol, so we put regulations around it.

Re: Dopamine: An AI platform for designing human behavior

#143

Post by tdaltonc » Thu, Nov 09, 2017, 10:44 PM UTC

013a wrote: post #142

mmayberry wrote: post #135

Earlier quoted context omitted.

you are correct. there is nothing wrong with giving humans the tools they want/need to be the better versions of themselves.

This is such a fine gray line that its impossible to conclusively categorize something as evil vs good. For example: If I decide I want to exercise more, maybe I would appreciate an app which helps me become addicted to exercise. That might be good for me. However, what if the app pushes me to exercise too much, and I begin to experience health problems associated with destroying muscles? Or, what if the app is forme…

The fact that there is a grey zone does not mean that there aren't also black and white zones. It more interesting to think about the grey zones, but it's easier to get things done in the back and white zones.

Re: Dopamine: An AI platform for designing human behavior

#144

Post by 013a » Fri, Nov 10, 2017, 5:35 AM UTC

tdaltonc wrote: post #143

013a wrote: post #142

Earlier quoted context omitted.

This is such a fine gray line that its impossible to conclusively categorize something as evil vs good. For example: If I decide I want to exercise more, maybe I would appreciate an app which helps me become addicted to exercise. That might be good for me. However, what if the app pushes me to exercise too much, and I begin to experience health problems associated with destroying muscles? Or, what if the app is forme…

The fact that there is a grey zone does not mean that there aren't also black and white zones. It more interesting to think about the grey zones, but it's easier to get things done in the back and white zones.

That's true, but I don't think a general purpose API for improving the addictiveness of any application exists in one part of the spectrum. Anyone can use this.

If we look at something like the activity circles on the Apple Watch; that's safe enough in my mind to be pretty well in the white area.

Re: Dopamine: An AI platform for designing human behavior

#145

Post by yters » Fri, Nov 10, 2017, 10:06 PM UTC

dredmorbius wrote: post #93

yters wrote: post #63

It's what TV advertising is all about. Why aren't we complaining about that?

Examine your premises. We are. And have been for a while now. https://en.m.wikipedia.org/wiki/Four_arguments_for_the_elimi...

Flipping out about a company that claims to modify behavior seems a little odd when they most likely consume hours upon hours of behavior modifying TV ads every single week. If behavior modification is the real concern, the majority of the angst should be directed at TV ads.