Hacker News Reader

Comment by mgraczyk | Hacker News Reader

Interesting to see posts like this where people speculate on specific factual things, but they don't have any information about the specific factual thing.

This post is factually wrong about how Facebook works and what the goals are. Facebook does not "uses every nasty trick it can to do one thing: keep you scrolling".

I worked on the feed algorithm for Instagram and know how the Facebook feed algorithm worked in 2020. I read all the code. The #1 goal of the Facebook feed algorithm is to get users to post more. The #2 goal is to get you to engage in interactions like commenting and liking. Facebook cares about how much you scroll and would like you to do more of it, but posting and interacting is way more important. Facebook optimizes for more posting at the measured cost of less scrolling.

Instagram does the same thing but at the time was less successful at it.

View on HN

Is it possible that Facebook's feed is not as good at that supposed goal as it is at getting people to scroll endlessly?

I've stopped using Facebook and Instagram in the last few years, so I can only speak from observing friends and family on those platforms... but I mostly left them because my friends and family didn't post. Once my feed became 90% ads and reshared posts from strangers, I left.

Maybe your definition of "posting and interacting" is different from the author's.

And on the subject of "every nasty trick [to] keep you scrolling": that includes, for instance, making it incredibly difficult to delete your account. Seriously, give it a try. You can't navigate to the delete page through the UI -- the only way to get there is to somehow find a link to the delete page URL. Facebook repeatedly asks you if you're "absolutely sure" you want to leave. They show you a slideshow of friends and family who "won't be able to contact you" if you leave. Then they make you wait 30 days (with no login attempts) to actually delete the account.

Seems pretty nasty to me.

> I mostly left them because my friends and family didn't post.

Yeah this is exactly the problem, and Facebook understands it. They want people posting more so that their friends will see their posts and stick around. Facebook cares about this more than scrolling because people stick around more in the long term if their friends and family are posting.

> Seems pretty nasty to me.

Yeah maybe, I've never tried. What I meant is that Facebook could do many things to get people to spend more time scrolling, but they don't do most because it would harm the user experience. For example Facebook was reluctant to add unconnected content on Feed, but eventually did. If they only cared about "keeping you scrolling" they would have done it 10 years ago instead of ~1 year ago.

I was on Feed from 2017 until 2020. Every year some minor sub-team would show off their plans for an unconnected content unit. It never shipped until Zuck began to understand that the ship had already sailed with Gen-Z on to TT (which he or a deputy refused to buy when he had the chance).

Everything Meta is doing today can be explained by Zuck's frustration with not owning a mobile OS (VR) and knowing that he has already lost the new generation (Reels and Rooms).

Not only did they mostly stop posting, but Facebook doesn't show the posts they make, giving me the sense that there's nobody to look at such a post

Great info, thank you for sharing.

Lets rewrite "keep you scrolling" as an inexact colloquial English expression that parses down to "spend time on site", and what they really mean to express: "form the daily habit".

Whether it is passively watching hands free a la TikTok or being prompted to press a like button, post text, or smoke a cigarette and regardless of the trigger - so long as you don't forget the thing exists and come back to it every day profits go up.

EDIT: No

Yeah, people attack Meta engineers for no reason. I wish you nothing but the best.

BTW, here is a cool game for you to take your mind off things, please enjoy thoroughly:

https://www.decisionproblem.com/paperclips/

Edit: please note before above reply was edited to 'No' it literally said that what I wrote is 'pretty much accurate'.

I don't work there anymore, and as you can probably tell from the stock ticker, Facebook is much closer to myspace than a universal paperclip machine.

It's just been a fad to be concerned about it. That's basically passed now, and the next few fads are already pretty mature.

MySpace peaked at 75 million users, currently Facebook is approaching 3 billion and still growing (just slower). Of course I am counting humans and you are measuring by stock ticker.

But hey at least it didn't drink the blood of all humanity, what's done is done. We should move on, you are correct.

Although, on second thought, it did spur those new next fads and set a precedent. And they do seem to be improving on the original in habit forming. It is also attempting to generate a new one to ride to a higher water mark and to be completely honest I'm not sure all those posts and things it prompted people to do en masse were beneficial to society overall..

Maybe there's some sort of lesson to be gleaned here.

> drink the blood of all humanity

Building a product that people for less time per day on average than TikTok, Youtube, Snapchat is "drinking blood"?

I think you should take a step back and evaluate if it's really as bad as you think it is.

> Building a product that people for less time per day on average than TikTok, Youtube, Snapchat

Firstly, that is despite your best efforts and not for lack of trying.

Secondly, as we established, whether it is 5 minutes or 5 hours, the key is the formation of the habitual use.

Thirdly, the 'drinking the blood of humanity' is about the Paperclips game. Even if you've played it before here is that link again to rejigger your memory: https://www.decisionproblem.com/paperclips/

Please do make sure to play it all the way through to understand the reference. A bit of a time commitment but really not as bad as you think it is.

Having a goal of using the product we make more isn't inherently evil, so why would anyone write about it that way.

> Whether it is passively watching hands free a la TikTok or being prompted to press a like button, post text, or smoke a cigarette and regardless of the trigger - so long as you don't forget the thing exists and come back to it every day profits go up.

You're just describing the basics of business but framing it to be nefarious? People who make a product, generally want it to be used as much as possible. It's why newspapers have crosswords and tv shows have cliffhangers. It's a problem when you want it to happen over the wellbeing of the users, so unless you think getting people to communicate with their friends and families is as bad as smoking, it doesn't seem to be the same thing.

It reminds me of that Key and Peele sketch where Peele described robbing a bank by working for them every day for 40 years

> It's a problem when you want it to happen over the wellbeing of the users, so unless you think getting people to communicate with their friends and families is as bad as smoking, it doesn't seem to be the same thing.

Facebook thinks its pretty bad: https://www.theverge.com/2021/9/29/22701445/facebook-instagr...

I don't know that it is worse than cancer. Lets hope not. Strange yardstick.

High level:

"We know Facebook executives believe that the company has positive overall benefits for the world"

And it was you that originally brought up cigarettes

“We know Facebook executives believe that the company has positive overall benefits for the world, and we also know that they are meticulous students of their own data,” Newton wrote on Tuesday. “It’s hard to understand why, if the data is so positive, Facebook is often so reluctant to share it.”

Please don't take things out of context. The authors really actually are claiming that Facebook knows it is harmful and knew all along - disagree with them if you want but don't rewrite their position.

> People who make a product, generally want it to be used as much as possible.

It's true for things that have a recurring revenue, and most of the time it's at the expense of the actual purpose of the product.

We take it for granted with social media, but imagine an email client doing the same and adding some random content between emails to keep you more engaged. Would you use it?

I worked on the same teams as you in nearly overlapping time periods and I add support to what you say.

FB is (or was, before I quit) driven by getting contributing users to contribute more.

We knew a million "tricks" to get people to be hooked on scrolling or watching ads. We did not use them, because we had already done a million experiments to know that people hate that shit, and quit. Interactions with friends is what drives use of social media, as it should be.

The trouble is, Facebook is so deep underwater PR-wise and user-content-quality-wise (in America at least, India still loves it), that subconscious evaluations of which app to open are facing very strong headwinds now.

> The trouble is, Facebook is so deep underwater PR-wise and user-content-quality-wise (in America at least, India still loves it), that subconscious evaluations of which app to open are facing very strong headwinds now.

Yes, the trouble is that ordinary people are making incorrect subconscious evaluations based on bad PR, such on the nose verbiage from you.

Is the app bad? No.

The app was foiled by that dastardly subconscious mind.

Most people (the layman discussing and planning to regulate companies like FB) consider interacting as part of scrolling, you’re being pedantic and “actually”ing here. The fact you worked on these algorithms is hilariously telling.

No, they are different because of how the UI works (also different on TikTok).

On facebook, you have to "enter into" a comment thread. Doing so makes it harder to scroll and causes you to consume less content. On Instagram this actually loads a new page and you can't scroll any more.

"Entering into" a comment section is bad for scrolling, it decreases ads consumption in the short term (per session) and the medium term (time scales of weeks or months). It increases ads consumption in the long term (years) very indirectly, because people enjoy the product more.

Even though it decreases "scrolling", Facebook attempts to cause people to "enter into" the comment section. This isn't an "achtually point, it's a real difference in how laymen would want to regulate companies like FB.

Now there are other problems with what I described, for example people might "enter into" comments to engage in political arguments or become polarized, but that's a different topic that Facebook also spends a lot of money attempting to solve.

This is not an important aspect of the discussion at all dude. The layman uses “scrolling” the way tech marketers use engagement and everyone knows this. No one cares about this attempt to strawman the actual problem here by yet again assuring they masses they’re too dumb to understand.

This is why these tech firms will be heavily regulated to death. They still don’t get that these minor nuances are things only they care about thus it’s passed time to let them figure this out on their own. They’ve talked passed the average citizen for long enough.

No, I disagree.

"Scrolling" means something like, "spending time on the app".

Facebook accepts less time spent on the app in the long term for more posting.

It's not hard to see how the author got there when FB and especially IG feeds are absolutely nothing but

"videos you just can't stop watching"

It's a huge turnoff for me and I close app as soon as I realize it's happening, meaning I use the app for about 5 seconds total each week now.

Maybe that's not what they were trying to do but it sure is what those platforms have become and does not bode well for Meta's future.

Thanks for the facts.

I think the basic premise of the article seems sound and unsurprising -- facebook users are the product sold to advertizers.

> That’s the lie that gets you in the door, and then Facebook uses every nasty trick it can to do one thing: keep you scrolling. As long as you’re scrolling you’re looking at Facebook ads.

What your comment suggests to me is that getting users to post and comment is probably instrumental to getting them to view more ads? More content, more scrolling, more ads?

Or is that mistaken as well based on your experience?

Yes in the long term. People who post are more likely to be retained for a long time for a lot of reasons, partially because they are more likely to receive feedback on their posts and the feedback keeps them engaged.

The ultimate long term goal of Facebook is to get everyone to use VR. The shorter term goal (3-5 years) is to get people to keep coming back to Facebook and enjoying it.

Lets assume for a second that Facebook's only goal is to maximise global scrolling, (measured in metres, say). They obviously need content for people to scroll through. If I post something compelling, potentially all my friends will scroll an extra 20cm. If I have 300 friends, that's a total of 60 metres extra scrolling in the world. More than I alone am likely to scroll in a day.

I would think that one thing FB could do to encourage them to post more is show how many people have seen something you post. Right now 50 people could see your post and all you know is that three people liked it and one commented. I think people would be encouraged to post more when they know they have an audience (assuming they actually do have an audience).

You might be overfixating on “keep scrolling”.

So the goal was/is to get people to engage more. So? Is the goal to keep people engaged for ad’s sake or not? Or is it to foster community and togetherness.

The thesis is about engagement with ads. Not whether you scroll or not.

Yes they are fixating on that because they’re still convinced society is wrong here, not them.

> to get you to engage in interactions

> like commenting and liking

But also, like blocking.

So if you block someone for posting something you really hate, you will see more posts that are like that post you really hated.

No, the opposite is actually true (I worked on things like this, it works)

I filter out the ads - so they definitely fail to optimize showing ads for me.

Which ad filters block facebook ads? IIRC a few years ago they had a surprisingly robust ad-blocker-blocker technique, using a crazy stacked div strategy to hide ad text.

FBP extension. Very feature-rich

commenting and liking or disliking.

Reply on news.ycombinator.com