Designing Social Systems Without Causing Depression and War
medium.com
medium.com
I guess when you only have a hammer everything looks like a nail.
Not reading admittedly boring and often inane philosophical books and instead trying to build your philosophy from first principles... tends to lead to whatever philosophy was created 50 years ago and has now permeated culture. Like how for a while, every geek I knew trying to re-invent moral philosophy ended up at utilitarianism, only to find out that the good counter-arguments had been made a long time ago; these days, people thinking about the internet, social pressure, and the nature of reality seem to be slowly re-inventing Foucault and Baudrillard without daring to touch the confusing French texts.
And kudos for overall quality post.
Currently so much rides on mere survival that the only effective objective function we have to consider our actions is that of money. Either the capitalist system needs to be diversified to encourage multi-objective optimization wrt quality of life across all lifeforms, or it needs to be dismantled and another form of cooperation that directly incentivizes helping your fellow human/animal survive should be installed.
Think about it like this: rather than writing some complicated optimization algorithm that sometimes climbs hills and sometimes descends them according to some arbitrary judgment scheme, we redesign the objective function so that all ethical behavior lies along stable manifolds. We can effectively hack human psychology by exploiting its predispositions to incentivize ultimately good behavior.
This "good behavior" doesn't even necessarily need to be complex or controversial: something as simple as the Golden Rule would likely be sufficient.
E: to add that I recently attended a contemporary art exhibition where one of the exhibits was the introduction of a new cryptocurrency called Kineticoin[1], which supposedly awards artistic contributions to society. How exactly to distribute this value is up for debate, but it is a sorely needed one, IMO.
Who gets to decide what is "Good" and "Ethical" behavior ?
What exactly would be the "incentive" "Good" and "Ethical" behavior. ? You say that capitalism can be dismantled in favor of another form cooperation that directly incentivizes helping your fellow human survive. Is this not in direct contradiction the base animal instinct of self survival. The incentive would have to be huge in order to override this base instinct.
You suggest Golden Rule as sufficient way to define "good behavior". But Golden Rule itself has shortcomings and criticism (https://en.wikipedia.org/wiki/Golden_Rule#Criticism). Do you have any sufficient argument against these criticism.
Sometimes reformulated as "An it harm none, do as ye list."
A suicidal person can say "I don't want other people to save me from dying therefore I will not stop others from dying."
A rich person can say "I do not want charity from others therefore I will not give charity to others."
Why should any one person get to decide? Right now we have given the power over that decision to the worst people.
I think a good first step is to recognise when someone has the power to incentivise behaviour and see whether they are using that power to incentivise negative behaviour. This is currently the case with most of the powerful institutions in our capitalist society. We do not live in a world of scarce resources. In most places there is more than enough food to go around, more than enough housing for everyone to be housed. We are forced into "survival mode" by the fact that wealthy people and institutions are hoarding resources and preventing access to them so that they can exploit us for profit.
If we simply checked and limited the power of such institutions, allowing people the freedom to act in a less competitive manner, we would probably have more co-operation and a more ethical society. We've seen plenty of examples of this before.
What about the basic social animal instinct of co-operation and empathy?
This is why people who sacrifice their lives for others are exceptions rather than norm. Seeing somebody lying on tracks in front of an oncoming train, the common instinct of onlookers is not to jump but stay where they are.
1. Persuade people to choose to be kind.
2. Manipulate people so that they act in a kind way (or avoid acting in an unkind way), without necessarily wanting to be kind.
Option 1 means you need a good way to remove unkind people from your own experience (“free listening”). This is typically not popular with commercial social network services, because this lets you ignore adverts too.
Option 2 reduces this need, but leads to the problem of who is “allowed” to do the manipulation, and who decides this?
When communicating in person, most people have years of learned behaviour (learned from millennia of social norms) nudging them towards acting kindly (or at least not actively unkind), because in person there are immediate, unavoidable social consequences. These consequences are absent when communicating remotely, such as via the internet. I'll let you decide for yourself whether this counts as option 1 or option 2.
To your criticism of option two, I think this is the fundamental flaw in the goal of engineered social systems. There are groups which have radically different ideas about what is "kind". Extreme religious fanatics would say it's unkind to let someone "live in sin", so extreme measures are considered kind for the spiritual well-being of the individual and the community. Secular western thought is more interested in material well-being, so it would say kindness is the promotion of the material well-being of others, generally without passing judgment on how they choose to live (except insofar as it violates others' material well-being).
So the decision of who decides what's kind can lead to completely different interpretations of what kindness is and how the social system behaves.
So what I like about the system, is that (for experienced users) down voting is allowed and that a down voted comment is getting less attention (less karma damage, less reader damage) and that is getting faded out (so while reading comments you can quickly skip the bad comments).
That way, you don't have to be afraid of writing a bad comment accidentally, because the system takes care of it, but when you write one you can see that that less people are reading it compared to a good one. I just wish down voting would require writing an answer. Nothing is more frustrating than a down voted comment when you do not know why people didn't like it.
Over the past years, I found that other readers like comments which tell them something new and insightful about the topic. Links for further reading and as references for opinions are also appreciated.
You very quickly get hive minded behavior: similar beliefs get upvoted, contrary ones get hidden, so soon enough you've got an echo chamber, powered by worthless Internet points.
It's a very basic hack on our psychology - we like points, but maybe it's not alltogether good. It works ok on HN because it's a decent community with moderation intended to keep the discussion good. But Twitter and most subreddits are horrible abuses of this and I'm not sure people would spend so much time arguing on the Internet if they simply weren't rewarded so directly for doing so.
I think the Slashdot system was an interesting one, it had also labels, so there was a difference between "+5 Insightful" and "+5 Funny". It also provided some feedback while downvoting. I think having moderation labels instead of plain points would help moderation.
Actually, users could then customize point cost for each of the labels, so they could see the page differently.
There are some great breakdowns of Reddit comment karma by time after post. The direction is obvious - earlier comments score better - but the exponential falloff is still a sight to behold. Any view that isn't outrageous enough to downvote heavily can win just by arriving first, being seen most, and being upvoted by commenters trying to increase their own engagement. (HN removes 'downvote' when people reply to you. I wonder if it should discount upvotes on comments you reply to?)
It's a less-trolled system than the open queue of a messageboard, and more conversational than the obscure branching pattern of Twitter or Tumblr. But it's not much good for comprehensive, noninflammatory discussion.
I would grudgingly agree, but found it funny that the following example sat on the front page the same day as your comment, and the second post exceeded the original's points & comments!
It is possible to detect and block Chrome headless | https://news.ycombinator.com/item?id=16175646 (currently 146 points, 105 comments)
9 hours later: It is not possible to detect and block Chrome headless | https://news.ycombinator.com/item?id=16179181 (currently 341 points, 158 comments)
A popular opinion doesn't make other opinions invalid and I think that's a mistake many people read into the situation because they're too emotionally invested. On top of that it completely ignores the countless variables that go into a top comment on a popular forum.
The best advice I can give is either 1. Don't take the unpopularity of your contributions so seriously or 2. take it so seriously that you're willing to understand how to game the system to MAKE your "objectively correct" opinion popular.
It does not bother me that unpopular stuff is downvoted (who really cares?), but it's a real dynamic. You said it yourself, if you want to craft a comment to be popular, there's a way to do that; it's not especially hard to figure out what to say if you spend time in various forums.
Maybe the issue with the upvoting system is it gives more credence to a baseless comment having value or being correct, when it's about popularity. From a social perspective it seems like a really great way to spread misinformation. Why would you not care about that?
In my opinion there is a system of homeostasis that exists in internet comments. When a popular wrong opinion gets large enough it is a huge low hanging fruit. There is a race to be the first one to debunk it and cash in all the delicious "well, actually" karma. Go look at the top comment on any given hacker news thread. I wouldn't be surprised to find an overwhelming number of "well, actually" child comments immediately below them.
Furthermore, popular anonymous comments are nice but the real world still operates on reputation based media. A popular comment might inspire a reporter at a media source of record to write an article about it but you can bet their first order of business is to cut out the rhetoric and make the opinion seem as objective as possible. In many cases once you cut out the rhetoric of a baseless popular opinion there is no article, so on a larger social scale the popular comment basically dead in the water. A firstname-lastname reporter is thinking in terms of a career, not upvotes. That's why when you read the sources of a non-fiction book you see The New York Times and not The Slashdot Comment Section.
* Also I just want to clarify that the "someone" in my original reply to your comment was not a veiled accusation, I literally meant someone (anyone). It was not a passive aggressive attempt suggest that you are getting too emotionally involved with your comments. I merely meant that a lot of people (myself included) get too caught up in thinking that 1. they are objectively right and 2. lack of recognition of this fact must mean that the system is broken.
Is it possible to have a forum on the internet with hundreds of active users that doesn't leverage a populist model (upvotes/favorites/likes)? Even places where the comments aren't ranked by the users they ARE ranked by activity (bumps/views) behind the scenes.
It seems like it would be very difficult to decouple "populism" from any large public forum worth reading so I'm skeptical of this criticism of HN. I would argue that it's disingenuous to accuse HN of mere populism without you defining your terms because even academics are skeptical of any one defintion. In many ways your use of "populist" is nothing more than a slur.
I am simply stating that "good" and "populist" are two different things. IMO, HN incentivize having popular opinions. Saying "HN incentivizes good behavior" comes with an underlying assumption that "populist" is "good" which I was trying to point out via my comment.
> I'm not sure if "populist" is as pejorative as you think it is when describing HN.
I am not using "populist" in perjorative sense.
> Is it possible to have a forum on the internet with hundreds of active users that doesn't leverage a populist model (upvotes/favorites/likes)? Even places where the comments aren't ranked by the users they ARE ranked by activity (bumps/views) behind the scenes.
Nearly all internet forums I have used leverage the populist model. However, I think this is more due to the designers of these forums. The designers of these forums want to show the most "relevant content" and their definition of "relevant content" seems to be the content with which most people agree. I believe one could experiment with different definition of "relevant content" to create different forums.
I personally would like to see a public indication of the ratio of upvotes to downvotes, and maybe comments to flags. Without this or some other system of checks & balances (even just a total number of submissions & comments on the user account page), completely anonymous moderation by HN users passing by will continue running the site. (Users on mobile devices who choose not to comment because it's a pain to include links to sources may deserve some consideration.)
The permanence of the site adds some unique properties as well: are there any other sites that lock edits after 1 hour or prevent deletion once there is a reply (or even any that attempt to accomplish anything even close to similar, whatever the means)?
This is feeling similar - that we need or want our virtual environment to help us - in practise sessions as he seems to say, or in advice
It seems we want science, but at the level of our personal, everyday, choices. we want to trust our computing environments have our best interests at heart.
and that kind of means no commercial interests can be involved.
If every person chooses for the better response and action given the a situation how can a individual be unique?
If all husbands reply to a bum in the exactly same way why do you really need a husband? Failing is a big part of being human, sometimes that can cause harm but failing and learning is one of ours core functions.
I rather be good than be you?
(Sorry if i made my post a bit 'agressive', its not my intent, english is not my main language and i still have to learn how to write properly)
I do agree with you that I'd rather make my own mistakes, than try to be a perfect partner by using an app. The idea is definitely interesting, but scary at the same time. It sounds like something out of The Black Mirror show :)
P.S. your English is great, nothing to worry about!
I'm wary of trying to reverse-engineer human behaviour, especially as it applies to personal relationships.
Besides, who decides what's the “best” course of action here? Who decides what the goal is — length of relationship? Happiness? (How do you measure that? Pleasure endorphins? Well then the correct response is you should stop talking to your partner, and do some exercise and drugs!)
In practice, the person who decides is the person in Google's office at 17:00 on a Friday.
(This is why driverless cars are hellishly scary, by the way. Any car that can be controlled by computer should have to use only publicly-auditable software — i.e. everything must be open-source and all builds must be fully reproducible. At least then you can theoretically be certain you know who your car will decide to save in an emergency, and why. How long before `git blame` is invoked in court in a murder case?)
Also, I don't think that asking for advice removes agency. You still have to think if the advice is applicable to your unique situation, how can it be modified to fit better, and which of several conflicting pieces of advice has more merit. An advice from Siri on any life situation is nigh useless without a deeper understanding of the situation, a Siri can't have it unless she lived a big chunk of your life. So, armed with whatever advice, you still remain ultimately responsible.
And for your second argument, it’s just to easy to follow the computers “orders”, a lot of people would just go for the order because it’s already justified, something along the lines of “I would do different, but the machine told me that everyone that on in that way get something wrong”.
I agree that this type of software would be good when one needs advices, but it would have to be complete trustworthy, giving wrong advices to a suicidal person by example can be devastating (and in this case not even thinking about the manipulation that would be possible by the corporation).