Hacker News Reader

Comment by zestyping | Hacker News Reader

This op-ed argues the question "Should Facebook Be Cut Down to Size?" but doesn't really address whether it can, or how.

Can it? What would work? How would you do it, if you could?

View on HN

It can.

The root cause of a lot of problems Facebook creates is the like count next to every utterances, article or behaviour on the net. This has to be regulated much more than private data/friend graphs/location data/what you view/search for etc etc.

Psychologists will tell you the most effective way to influence behaviour of an individual is to show people what the herd around them is thinking, saying or doing. This is how society has always created "social norms".

If people around you don't smoke, you are less likely to be a smoker. Teen girls who see other girls get pregnant are more likely to get pregnant. There are literally thousands of examples that can be cited about how seriously we take this signal.

In a world with too much info that no one is trained to handle or are too busy or distracted to handle, we look at what what the majority of people around us are doing and do that blindly. This is signal influences behaviour. Both good and bad.

Facebook has fucked that signal up for society. In every sphere of life. And we are currently dealing with the fallout. The signal is being delievered too fast and too much imho. And this speed is highly overrated in its value to society.

Whether these counts should be displayed, to who, how quickly, next to what type of content, for different communities, education levels etc needs to be debated and regulated.

> Whether these counts should be displayed, to who, how quickly, next to what type of content, for different communities, education levels etc needs to be debated and regulated.

Yeah, but wouldn't that apply to HN too? What if they decide that we're seeing too many different articles posted on HN? That we might get stressed by the amount of new computer languages we hear about, or depressed by the number of Show HN posts ...

Most definitely it does effect the HN users. Every site that uses the upvotes mechanism as a signal to the group about what they should think or focus on is producing both positive and negative consequences. It's used not because the good outweighs the bad but because it's implemented by programmers who find it easy to implement. If you asked a social psycologist/socialogist/ecologist or an educator with experience in building learning environments a very different system would be in place. We need these kind of people influencing these feedback loop not just the well intentioned but misguided programmer.

I have bought this up multiple times before. There are a lot of impressionable minds around here being exposed to lot of half baked/frustrated/misguided thoughts. Many times the upvotes mechanism just amplifies and reinforces these thoughts.

> I have bought this up multiple times before. There are a lot of impressionable minds around here being exposed to lot of half baked/frustrated/misguided thoughts. Many times the upvotes mechanism just amplifies and reinforces these thoughts.

So people should only be able to see vetted information? Who decides whats not a "half baked/frustrated/misguided thoughts"? Seems like a slippery slope.

The programmers who implement these systems don't even know that their upvotes mechanisms can have a down side. Only after they acknowledge it as an incomplete and incorrect social signal and that they are out of their depth in controlling the unintended effects can the debate begin.

Good point. While we're at it, let's get rid of Amazon and Netflix recommendations, Google search results ordered by quality and relevance, and email spam filtering.

So you think the like counts don't have negative effects on people's beliefs and behaviours? Please read Nudge by Richard Thaler to understand how easily misguided the herd is when following the wrong signals.

You don’t see a difference between rating products for consumption, and rating people and their behavior? Spooky.

I don't think it can. The AT&T breakup worked, because even then, if you had a phone number, you could contact anyone else with a phone number, even if they weren't an AT&T customer. Facebook, and the other social media silos, are built with anti-compatibility as an explicit goal. That, coupled with Metcalfe's law, means that any upstart will never be able to directly compete with Facebook (or another social network) on it's ground.

How about coming up with some standards to allow federated social networking, and requiring every major social network to follow this standard?

perhaps federated protocols for all social media.

But a post on Facebook looks different from a post on Twitter. Could they really share? Could you write a regulation where it makes sense that they're different?

Maybe you could. Twitter and Tumblr seem to be about the same modulo the length. Isn't the advantage of Twitter its shortness? Maybe we'd discover what the advantage of Twitter is.

And perhaps Facebook is the same just with better networking features. Maybe we should try to write federated Facebook and Twitter clones to prove it can be done.

The only difference is that a post on Twitter has artificially limited size. This is a restriction imposed by Twitter and as such could be handled by them - for example by showing a short summary + a link to full message for posts that were imported from other networks.

Requiring reasonably-formatted downloads of "friends" lists would go a long way toward making switching networks frictionless.

Whether or not such a requirement would be good for privacy is an important question.

Obtaining your list of friends is not the factor that makes switching social networks unattractive, it is the friends themselves.

The list doesn't help if your friends aren't on the other network. HN loves to come up with technical solutions to the problem of Facebook when it largely isn't a technical problem, it is a social one.

It makes a huge difference whether you can pre-populate a new network with a list of your friends, as it makes your friends' discovery possible.

Imagine, which Google+ launched, how things might have gone if 5% of Facebook users could have moved their own social graphs to G+ with two clicks of a mouse. If everyone had 100 friends, and those friends were uniformly distributed (all stark simplifying assumptions), then a huge fraction of the global social graph could have been moved to the new network by a small fraction of the users.

But you can't mandate a business model/technology. SS7 and IP aren't legislated worldwide, they're in place because it is in the best interest of telecom companies to be able to interconnect.

Social media has no such incentive, so how could a regulated federated standard work? Even if it could, who would pay for / provide the development? The social media companies certainly wouldn't (https://xkcd.com/927/), and governments wouldn't for largely the same reason.

I would think this would be _really_ hard. So many of the simple features that make up a social network are incredibly complex on the backend. Things like hashtags, searches, recommended friends. Are giant specially indexed and formatted databases. If all social networks did was send messages that would be easy but how do you do a newsfeed amalgamating multiple companies? Does everyone have to use the same algorithm for newsfeeds? Who’s comments come up first, what does search look like? How about spam control, are you allowed to block posts from an unregulated social network? Do you all have to have the same content guidelines and ToS?

Do you think it might work to tell Facebook that in a year's time they'll be broken up into (for example) 5 separate companies, and their users will be assigned at random to one of those parts but must also be allowed to move between them at will?

That gives Facebook engineers a year to design and implement a federated API that would work seamlessly. If reasonable and non-discriminatory licensing was required on the API standards, then possibly others could interoperate with it too. Hopefully the mini-Facebooks could start to differentiate themselves through features, price, privacy, etc...

I'm not completely convinced a federated protocol in and of itself will solve the entirety of the problem. I think we need some blockchain and smart contracts, people need to actually control their own data on these services. For instance, if facebook created an open source wallet, your data could sit behind that on your own device.

Needless to say, I think even if congress ordered facebook to breakup, it would be cheaper and easier for Zuckerberg to just move everything off shore and thumb his nose at a relatively impotent body. Zuck could also just lobby his ass off and probably end up writing the laws for the congressman.. Unfortunately, business writing the law for itself is pretty common in the USA, I'm sure Zuck knows that.

Making WhatsApp and Instagram separate companies would be an easy start. Then you could also separate advertising from social networking, e.g. let advertisers compete on FB. Would it work as in 'would all components survive?'. That we would find out soon enough.

> How would you do it, if you could?

Easy starting point: Facebook, Instagram, WhatsApp.

the obvious place to start splitting is to make instagram and whatsapp independent again. it's crazy to me that facebook was able to purchase it's biggest (and in many regions, only) competitors in the social media and messaging markets.

It's actually really simple: split off Instagram.

Reply on news.ycombinator.com