Random Thoughts – Randocity!

Why I stopped using Twitter

Posted in botch, business, california by commorancy on November 25, 2022

a woman in red scarf holding a megaphone

Based on my recent article, Is the Demise of Twitter imminent?, I have outlined the reasons why I believe Twitter is very close to closing down entirely. While that is a reason not to use the platform, it isn’t my primary reason for leaving Twitter. Twitter has a lot more wrong with it than potential closure. Let’s explore.

Content Moderation and Trust

Let’s jump right into the heart of the reason why Twitter is in serious jeopardy. Any social network that offers User Generated Content (UGC) is at risk if the operators of the site are unwilling to handle that UGC appropriately.

Terms of Service (TOS) agreements and Acceptable Use Policies (AUP) exist to protect the site from lawsuits. Meaning, so long as the site adheres to the terms laid out in their agreements, then the site is said to be doing its fiduciary responsibility to its users.

TOS and AUP agreements define what is considered acceptable conduct by anyone who uses the web site. Most such agreements lay out that conduct such as hate speech, harassing speech, bullying, threats of violence, death threats and any conduct which is considered illegal federally or locally is prohibited on the web site. The article I mentioned above also touches on this topic.

Whenever a site is created that publishes such user generated content on behalf of its users, a site must make sure that the speech remains within the confines of acceptable use. That means offering such mechanisms as user reporting features (allowing users to report offensive content), automated scanning of content to detect such infringing content and a team of content moderators to remove or suspend users who willfully break the rules.

Why do these agreements exist?

Trust. These agreements are in place to help users understand that Twitter is a safe and trustworthy space. As long as the agreements are upheld, then users can know that Twitter is looking out for them. Without such agreements or, more specifically, knowing the agreements aren’t being enforced, then the safety level of the site drops precipitously, along with the site’s level of trust.

Politics and AUP

Recently, too many people on Twitter are now seeing everything through the a political lens. Specifically, the right wingers are now seeing everything they say through a political lens of free speech.

Let’s understand first and foremost that First Amendment Free Speech DOES NOT apply to Twitter or any non-governmental organization operating a social network. It never has. The First Amendment only applies to Governmental organizations and staff. While your local county official cannot abridge your freedom of speech or freedom of press, Twitter can.

Further, let’s understand that terms of service (conduct) agreements are not built with politics in mind. They are built by lawyers who are paid to provide legal services to corporations. These agreements are not political leaning. These agreements apply to everyone using the services equally. Anyone who infringes the agreement is subject to disciplinary action… yes, ANYONE.

Right Wing Activists and Lying

Right wingers have been completely jumping on the bandwagon that somehow Twitter is selectively applying its rules only to right wing activists and not to left wing activists. That would be unfair application of terms of service, but it’s also a false statement. That kind of false rhetoric is now a staple with right leaning conservatives. They’re willing to lie about nearly anything and everything. Why would social media be an exception? It isn’t.

Twitter has applied its rules equally to all people who infringe, left, right or center. It doesn’t matter what your political beliefs, if you put forth infringing content, you’re suspended or banned.

Left wing activists have also been banned from the platform. Thus, this right wing falsehood is just that, a falsehood… like many others. Yet, they keep saying it with careless abandon as though saying it multiple times will somehow make it true. It doesn’t.

As of this moment, right wingers are completely out of control on Twitter… running afoul of Twitter’s rules without any disciplinary action by Twitter staff. That’s not to say left wingers aren’t out of control, because they are also. In fact, there are a lot of apolitical people on Twitter simply playing games with Twitter’s rules because Twitter isn’t enforcing them…. and here is the problem in a nutshell.

Rules, Chaos and Crowd Sourced Moderation

Rules exist to stem the chaos and enforce trust. Without enforcement of rules, a social media site is simply a cesspool without trust… and that’s exactly where Twitter sits right now.

If Twitter had been designed to allow thread creators to manage and moderate user comments within their created thread, like YouTube owners can moderate comments on videos, then Twitter would be in a much better place right now.

It would mean that I, as a Twitter user, could dump off comments from my thread that break not only Twitter’s rules, but my own personal rules of decorum. Unfortunately, Twitter doesn’t afford that level of content moderation to the thread creator. That means relying on Twitter’s now non-existent staff. Of course, when that staff doesn’t exist, there’s no one there to do the moderation work that’s needed.

If Twitter had moved to crowd based moderation, the platform would be in a much better place. It wouldn’t need nearly as much moderation staff as thread creators could simply remove comments from threads they own. If someone chimes in with an insensitive, inappropriate or problematic comment, then “Delete” and the comment is gone. No Twitter staff needed.

In fact, this is the way social media needs to operate now and in the future. Twitter still firmly believes that it is Twitter’s staff sole responsibility to moderate content. That’s not doable when you have perhaps billions of messages being sent daily. A company can’t grow its moderation team to scale to this number of messages. It is also an antiquated idea that should have been gone years ago. However, at the time of Twitter’s conception, crowd managed UGC wasn’t really commonplace. Partly that’s something that wasn’t being done, but partly it’s because Jack Dorsey’s team didn’t have the foresight to realize staff moderation of billions of small messages was not humanly scalable.

In recent years, crowd managed moderation has become not only more acceptable, it’s become commonplace and even important. YouTube has allowed this for quite some time. It allows the channel owner to remove any and all messages from its videos that the content creator deems problematic. It firmly puts the burden of content moderation on the creator. That’s also a completely acceptable situation.

Crowd Moderation

Wikipedia has completely proven that crowd moderation of content works. As a company, you can’t afford to hire the thousands of people needed to scout billions of messages all over the platform. Instead, it’s better to empower content creators to manage a much smaller number of messages.

Reporting inappropriate comments is still available, however. This allows staff the opportunity to jump in and manage inappropriate content if the content creator reports a comment.

However, conscientious creators should be willing to hold and moderate comments prior to allowing them to be published. With Twitter, publishing is instantaneous with no advanced moderation possible. Considering the sheer volume of messages on Twitter, it might be almost impossible to handle a single hold-queue style moderation system. With a spam filter, it may be possible to separate the wheat from the chaff into more easily manageable piles.

Trust, Quality and Moderation

Here’s something that Twitter has needed for a very, very long time. Twitter is chock full of bad actors. Any bad actors who consistently write bad comments of low or questionable quality would see their comment moved into the “junk” moderation pile for the content creator to manage and/or report.

Such a system would allow Twitter to offer up content moderation for all of its content creators. Enabling content moderation places moderation in the hands of the content creator using a hold queue. This halts many instant responses, but it ensures higher quality comments. Comments are then examined and filtered into trust and quality buckets. High quality comments from more trusted individuals get placed into the pile the creator manages first. Successively lower quality comments from lesser trusted people get moved into successively lower moderation piles.

Content creators can both move comments from one pile to another and they can mark commenters so that future comments get placed into specific piles all the way to a block which prevents the user from commenting at all.

For example, piles might be labeled as:

  • Instant Publish
  • Mostly-Trusted
  • Semi-Trusted
  • Untrusted
  • Untrusted Junk
  • Junk

These 6 piles are a good starting place. Instant publish is for your most trusted followers. You know that these followers can be completely trusted to instantly publish a high quality comment with no holds. No moderation is needed for fully trusted people. For people who are mostly trusted, these comments go into the mostly-trusted pile for moderation hold. These are people who are very close to getting instant publish, but you still need to hold their messages because you want to read the comment first.

All other piles are reviewed at the sole discretion of the content creator. If the content creator chooses not to look through the remaining piles, then the comments get purged after 7-30 days on hold.

How does a user become trusted?

Trust comes from both following and adding a new button labeled ‘trust’ along with an assigned level (1-6). Following someone only places someone into the Semi-Trusted pile. Meaning, you’ve followed them so you’re assigning them the default trust of level of 3. However, you haven’t completely trusted them. This means you’ll need to moderate comment content.

As a user gets more and more messages posted out of moderation, the user will automatically move up the ranks of trust, eventually reaching Instant Publish unless the content creator explicitly sets the user’s trust level.

User trust levels can also be managed by interactions with others. A content creator can enable “inherit trust averages” to new followers. This means that user’s trust level is calculated and inherited based on past interactions. If a user has consistent bad interactions, been reported a number of times, been blocked by many people and so on, these bad activities affect the user’s inherited trust level and the user’s trust level goes down. Instead of being assigned a default of level 3, the user might inherit a level of 5 or 6.

Note, being blocked by lower trust level users doesn’t influence a user’s inherited trust. Only people of higher trust levels who block them influence the inherited trust level. This stops bad seeds from gaming this system and attempting to lower a person’s trust level by creating hundreds of accounts and blocking someone of higher levels of trust. The only trust levels that impact a user’s inherited trust level when blocked is if the blocking user has a trust level above 2. That means bad seeds would need to work their hundreds of accounts up to level 2 before blocking people to reduce trust. Even then, any user attempting to game the trust system will automatically be banned.

Note that there are effectively two trust levels at play. There is the inherited trust level of the user themselves, which is gained by behaving correctly, producing high quality content and, in small amount, by having someone follow you. The second trust level is set by a content creator. Even if a person is inherited with a 90% trust level, if they follow someone and comment, the content creator can set that 90% trusted user down to level 6 if they choose. That moderation trust level only applies to the content creator, but doesn’t impact the follower’s inherited trust level… unless many high level trusted people all mark that user down.

Trust levels are the means by which the bad actors go to the bottom of the pile and good actors bubble to the top. To date, no social networks have instituted such a trust system. Instead, they have chosen to allow chaos to reign supreme instead of forcing users to learn behavioral norms when interacting on social networks. Enforcing behavioral norms is something social media desperately needs.

Trust Numbers

Implementing a trust numbering system would also add more control by users and content creators alike. Users who insist on being untrustworthy, to lie, to generally be toxic will see their trust numbers reduced. It doesn’t matter if it’s a celebrity or a nobody. Trust numbers are what people will judge. Like any score system, it can be used to allow users to auto-block and auto-ignore users who choose to have trust cores below a certain threshold. If a comment from a user with a trust level below 50 would appear on a timeline, a rule saying hide comments from users below trust level 50 would automatically weed out toxic comments.

More than this, if a user has a less than 50 trust score, a content creator can make a rule that prevents low score users from commenting at all. In effect, the trust score auto-blocks the user from comments. If the user wishes to make a comment, then they need to do the right things to raise their trust score. A trust scoring system is the only way for users and content creators to know that they can be safe on a platform like Twitter.

Chaos now reigns at Twitter

Because Elon Musk has decided to cut over half of Twitter’s staff, there’s really no one left to enforce much of anything on Twitter. In effect, Twitter is now overrun by untrustworthy, lying, conniving bad actors. It is these toxic people who don’t deserve to have any interactions at all. They are the absolute dregs of social media. These are toxic people you would never interact with in person, yet here they are on full display on Twitter.

Because Twitter has no moderation staff left to manage these bad seeds, the platform is overrun by people of bad intent. These are people who insist sowing seeds of chaos and doing as much damage as possible all with providing no value to the platform. Their comments are worthless, bordering on toxic and are sometimes even dangerous.

With no moderation team, there’s no one at Twitter who can review these comments for their toxicity, let alone do anything about it. Worse, Elon Musk is pushing a “new freer” Twitter, which simply doubles down on this level of toxicity all over the Twitter platform.

If Twitter were to introduce a trust and moderation system as described above, Twitter could forgo the moderation staff, instead letting content creators manage these bad seeds to push them off of the platform. Such a moderation system would also take a huge burden off of Twitter’s staff. Bad seeds would eventually disappear when they find their comments don’t get published. They also can’t claim Twitter is a fault because a content creator moderation system would mean people of all political persuasions would be kicking these bad seeds to the curb.

There’s really no other way for Twitter to manage such bad seeds other than a crowd managed moderation system like the above. Unfortunately, Twitter’s staff is dwindling at an astonishing rate, including the very software engineers needed to design and build such a system.

If Twitter wants to become a platform about trust and safety, it needs to institute a mechanism that enforces this philosphy, like the above content creator moderation system. Without such a system, Twitter remains chaos.

Toxic People

Toxic Symbol

Toxic people are everywhere, but it seems that social media like Twitter attracts them in droves. I don’t know why other than the anonymity that seems afforded. Suffice it to say that while Twitter was relatively toxic prior to Musk’s takeover, the content moderation staff took care of a lot of that toxicity through suspensions and banning.

Unfortunately, Musk seems to have reversed that stance and is now allowing (and even condoning) toxic people back into Twitter who were formerly removed. That means Twitter is now becoming even less of a safe and welcoming space than it formerly was. Toxicity now prevails. Toxicity is something no one needs in their life, least of all on Twitter. Toxic people are draining for all of the wrong reasons.

  • Toxic people waste your time — Toxic people ask you to do stuff for them while providing nothing in return. Even if you do spend the time providing what they request…
  • Toxic people always criticize you — Wasting time on someone toxic, they will turn that wasted time against you by arguing and criticizing what what you provided was not what they requested.
  • Toxic people spread negativity — Even after trying to talk to them to convince them, they will still turn it back around on you as a negative, as though you did something wrong. You didn’t.
  • Toxic people are jealous — The most likely reason they interacted with you in the first place is that they are jealous of what you have. In order to make themselves feel better, they will argue and downplay over whatever they are jealous… or they will try to make you feel jealous by claiming they have something that they don’t actually have.
  • Toxic people play the victim — Instead of accepting their own faults and failings, it’s always someone else who is to blame for them. If you happen to get in their way, you will become the victim over their having been victimized by you. That goes back to being jealous. If they are jealous over something, they will blame you for their being victimized by their own jealousy.
  • Toxic people are self-centered — This is a form of narcissism. How bad the narcissism is depends on them, not you. This means that not only are they likely to blame you for them being a victim, it all revolves around them, never around you. These people never see you as anything more than a punching bag to inflate their own ego.
  • Toxic people really don’t care — In other words, they argue with you because it inflates their ego, but honestly they don’t care about you or how you feel as long as it makes them feel better. It’s a form of manipulation.
  • Toxic people will manipulate you — This is another form of narcissism. It all ends up revolving around them. Most toxic people don’t care about your feelings at all. All they care about is getting whatever they want out of you. If that’s money or a ride or food, they’ll do or say whatever makes that a reality. On Twitter, you have to be cautious as money is really the only motivating factor. If Twitter enables money transfers, expect these toxic people to turn into scam artists.

Twitter currently enables, facilitates and now condones these toxic types of people on Twitter. Not only will they waste your time, they will attempt to play the victim game as though you caused them to be the victim. They will always claim that you are the one who is wrong and they are the one who is right. There is no middle ground, concession or compromise with toxic people. It’s always them and no one else.

If you feed into their garbage, you are likely the one to be harmed by them. Don’t allow it. As soon as you see someone like this, block them instantly. Don’t interact with them. If Twitter isn’t willing to handle toxic people, you have two choices, block and hope they don’t come back using another account or stop using Twitter.

Leaving Twitter

What Twitter currently means for sincere AUP-abiding content creators is increased effort to block toxic people, which actually does little to stop that user’s toxicity. They simply move to other victims to vomit their toxic rhetoric, with those users being forced to block them also. In other words, there’s nothing at all a standard user or content creator can do to stop toxic people from being toxic on Twitter (other than blocking that person for themselves). The best a legitimate person can do is block these toxic people for themselves alone, but that doesn’t make any impact on that toxic user’s account. Even reporting such an account today is likely to go ignored by Twitter. Musk appears to have no interest in holding rule breakers accountable.

A trust system would change this game. Meaning, users who insist on being toxic get to share in their consequences of being toxic. The more toxic they become, the more their account gets moved to the bottom. When the account gets down to a certain threshold, this allows Twitter to review these accounts for being a problem… thus requiring far less staff.

Unfortunately, Twitter has now placed this time suck burden onto each user to block, mute and dump users and to clean up the mess after. I don’t have time for that. Not only is that a complete waste of my time, I’m not being paid by Twitter to do it. It also means Twitter is not a safe or welcoming space. Spending my time managing my account only affects my account alone. It doesn’t in any way stop those toxic bad seeds from laying siege to other users on the platform. Since Twitter has no staff to manage these toxic bad seeds, Twitter is simply a chaotic cesspool of the lowest social media dregs all running amok in a quagmire of chaos. No one is safe from these toxic people.

If you’re looking for a safe and trusting space where you can feel like the social media site is looking out for you and your best interests, Twitter is not that place. Twitter has now become literally the worst, most toxic environment you could join right now, second up only to Facebook. Twitter doesn’t care about trust or safety or protecting you. They’re only interested in letting toxic social media users run roughshod all over everyone else.

For the reason of toxic users and Twitter actively choosing to be unsafe, I am off of Twitter. I simply cannot condone using a platform where the management is more interested in allowing chaos to rule over offering up appropriate safety measures for its users to use against toxic people.

Twitter’s Safety Rating

Safety: 1 out of 10
Toxicity: 10 out of 10
Recommendation: Avoid until Twitter closes or Musk figures it out

↩︎

Rant Time: Pinterest

Posted in botch, business, Random Thoughts, rant, reviews, social media by commorancy on June 30, 2019

pinterestPinterest is an image sharing platform using image ‘pins’, which should be interesting. After all, the word “interest” is in its name. You would think that before releasing a platform designed around relevance, the Pinterest team could actually design an engine capable of producing relevant and interesting images. NOT! Pinterest is one of the worst, if not THE worst platform, at displaying relevant ‘pins’ in your feed, not that Tumblr and Instagram are much better at this. Let’s explore.

Search Interests

One of the things that has vexed developers for a long time is how to show you stuff that’s actually interesting and, more importantly, relevant to you. Amazon and Google have done a decent, albeit not any anywhere near perfect, job of implementing such search heuristics, where the results actually offer some measure of interest and relevance to you based on the data they know about you.

This data collection, storage and mining issue is currently a point of privacy contention among many and is even in the news, but ‘search history’ is the primary means of showing you “stuff” that is actually of personal interest. The secondary method, which is less creepy and at least a bit more tolerable, is asking you directly for categories you’re interested in (i.e., sports, fashion, music, your age, single/married, kids, etc). Still, your search history actually contains the most relevant information about you as it’s recent and current. Unlike family relationships that can change (kids grow up, couples separate, graduate from college, move, get remarried, etc), search history implies a lot about your current situation and is way more up-to-date than explicitly given data that gets old even just a month or two after it’s given. Explicit offered data can even be based on lies, because some people roll that way.

As an example of recent search history, searching about baby related stuff (cribs, clothing, formula, diapers) might yield ads from Amazon, Target or Walmart selling baby goods. It only makes sense… and this is an example of ‘relevance targeting’. That is, targeting you with images or ads you have searched for in the recent past. Same for searching for wedding, bridal or other similar information. Same for searching for car buying. Search history is ‘in the now’ information that is clearly relevant to you “right now”. The “right now” portion of search relevance is key to a great relevance engine and to ad targeting.

Unfortunately, it doesn’t work if you share a computer with multiple people; for example, you might have a family of four or have roommates in your dorm. In cases like these, your daughter might have searched for Barbie dolls and now you have a bunch of irrelevant (to you) stuff related to Barbie or toys or kid related items. The search engine simply can’t recognize who is at the keyboard. It currently can only attribute search results to a specific computer. Until search engines can identify who is at the keyboard with each search (i.e., facial or voice recognition), engines must identify based only on the computer itself (a limited recognition system). This is the reason voice assistants like “Ok, Google”, Alexa, Siri and Cortana are so important. Unfortunately, I don’t believe these assistants yet identify the voice itself. They only recognize the words spoken and translate that into text search.

Search relevance definitely isn’t perfect much of the time and doesn’t work at all when using a shared device. Using a shared device, I do get why ‘relevant ad targeting’ doesn’t work. However, if your device is solely used by you, then relevance targeting should work perfectly… or at least as perfectly as today’s targeting algorithms allow. Yet, for Pinterest, it doesn’t.

Pinterest’s Targeting Engine

Why discuss the above? Let’s illustrate exactly how Amazon and Google work ad targeting relevance. If you’ve searched for “men’s clothing” in the recent past, then Google and Amazon will insert these kinds of items into your ‘feed’. A feed is basically a place on the screen where ‘Recommended for you’ stuff appears.

Pinterest doesn’t use a ‘Recommended’ area, instead choosing to intermix it all together in one immediate and immense jumbled mess. If any dictionary needs an example for ‘cluttered’, Pinterest certainly works.

[RANT ON]

Like bread falling butter side down, so this rant begins. Pinterest has one of the worst designed, most sloppy, most cluttered, most inaccurate relevance engines in existence. In fact, I don’t even think Pinterest has a relevance engine. They seem to vomit up all random irrelevant garbage into your Pinterest feed based on who knows what criteria.

Worse, they then attribute that random spewed garbage to being ‘Inspired by’ (a form of ‘Recommend’) to a board you’ve created. I’m sorry. Wait.. what, Pinterest? How does a picture of a baby in a carryall at all relate to men’s fitness? Seriously, I’m a single guy. I am not currently in a relationship. How does a picture of a baby at all interest me or, more specifically, how does that picture of a baby relate to body fitness? Clearly, a baby is not the definition of ‘fitness’. That is, unless Pinterest is actually trying to promote pedophilia?!?

Worse, I also see pictures of fat hipped women that claim to be ‘Inspired By’ a board on men’s bodybuilding and fitness (no women in that board at all). I also see women’s hair styles flooding my feed claiming to be ‘Inspired by’ a board on men’s underwear. I see pins of women in wedding dresses. I see pins of women’s high heeled shoes. I see women wearing random fingernail polish and acrylic nails. These are entirely fashion related and I have not a single board or pin devoted to women, women’s fashion or, indeed, women’s anything. Not a single board. How can you possibly claim attribution of these completely random images to any board in my account?

I’m not against any of these topics. If I want to see them, I’ll go search for them and look at whatever pins are there by searching. However, I DON’T want them in my Pinterest feed. These pins have no place there.

It Gets Worse

From here, Pinterest’s relevance goes into the toilet (literally… yes, bathroom cleaners are there too). I get that Pinterest might think a single guy might have some interest in looking at the female form dressed or coiffed nicely. But, even if that’s true (and in my case it isn’t, at least I don’t want it in my feed), Pinterest insists on throwing all manner of completely irrelevant garbage into my feed.

It’s not simply limited to arbitrary women’s fashion, oh no no no, my reader friends. Pinterest insists on throwing Arabic writing into my feed… a language I not only cannot read, I also have no interest in. I’m sure that whatever is written there is fascinating, too bad it’s wasted on me by throwing it into my feed, an English speaking person.

It gets worse. For at least six months (maybe longer), my feed was entirely littered with page after page of all manner of tattooed body parts… just the parts. These included ankles, wrists, shoulders, backs, butts and torsos. Sometimes the tattooed body part is so close to the camera, I can’t even identify where it is. Worse, the tattoos are downright fugly. They looked like someone had done it themselves DIY at home. A few were professionally done, but many were so horrendous, who would even consider putting such a thing on their body? Anyway, I have no tattoos, have no interest in getting tattoos and don’t want to see tattooed body parts in my feed. I hadn’t searched for tattoos, so Pinterest didn’t get this ‘idea’ from my search history.

These tattoo body part pins were literally clogging up my feed. Nearly every image in my feed was of a body part. I might understand seeing a little of these images occasionally. As I said, it didn’t come from search. However, while I did have a fitness board that incidentally contained some men with tattoos, they were there because of their physique, not because of their tattoo. Pinterest doesn’t get it. It only saw a tattoo and then insisted that I might have some interest in tattooing my body… thus flooding my feed with body part after body part with UGLY tattoos. A completely wrong assumption, I might add.

Assumptions are, in fact, the prerequisite to search relevance. Unfortunately, Pinterest’s assumption engine is entirely wrong nearly 100% of the time. Just because an image contains a tattoo on someone’s shoulder, you can’t assume that to mean I want to tattoo my body and need help by flooding my feed with tattooed body parts. Wrong assumption, wrong results… or as the older computer adage goes, “Garbage In, Garbage Out!”

Pinterest Janitor

Here’s where it turns REALLY ugly. To clean up my feed, I had to play janitor. First, I had to spent valuable time going into all of my boards and clearing out ALL pins that had ANY tattoos in the image. Just gone… out of there. That helped a little, but only a tiny amount. It only helped a little because Pinterest’s engine had already ‘learned’ this ‘interest’ based on an incorrect assumption. Unfortunately, ‘unlearning’ learned stuff can he incredibly difficult… and, in Pinterest’s case, it is! Second, I had to spent time going through each new “tattooed body part” pin appearing in my feed, then following that pin through to the original account who pinned it… and then, you guessed it, block the account. That all sounds easy enough, but because of the way Pinterest works, it’s actually quite time consuming jumping from page to page and waiting for Pinterest to refresh each super long, image laden page.

I spent the better part of a week opening pins, going into accounts and blocking account after account after account. Blocking the account is the only way not to see these pins in the future (well, sort of… this is actually broken, too and I’ll discuss this next).

You’d think that a platform like Pinterest could figure out a way to wholesale remove an interest category from a feed… but you’d be wrong. Nope, there is no way to remove an interest (or should I say, exclude non-interests) from the feed. The only way to remove an interest is to, one by one, block the accounts producing the pins. It’s the only way. Even then, new accounts spawn all the time leading to brand new pins of the same old stuff recycled back into my feed… requiring even more blocking. It’s a never ending janitorial cycle.

Now, you might be asking, “Why not click the … (ellipsis) menu on the pin and report it?” I tried that. It doesn’t work. Reporting the pin as spam does nothing. The pins continue to show up. The only way to stop a pin is to block the account who pinned it. Even then, blocking an account has limited ability to even stop the problem…

When Blocking Doesn’t Work

You might think, once again, that blocking an account would block all pins by that account. Again, you’d be wrong. The only thing that blocking an account does is block pins created directly by that account. If a different unblocked account repins one of a blocked account’s pins, it can still end up in my feed. Repins via unblocked accounts allow pins through from accounts that are blocked. It’s not the pin that’s blocked, it’s the account. This is a huge heuristic mistake for a platform like Pinterest.

Even then, blocking an account doesn’t take effect immediately (or sometimes even at all). Pins that are already in your feed stay in your feed, even after you’ve blocked an account. I’ve blocked accounts and for several hours after continued to see that account’s pins in my feed after refreshing multiple times. A block seems to take up to 24 hours to actually take effect fully. Even then, I’m not entirely certain that blocking does much good because of repinning. Repinning is Pinterest’s version of Twitter’s retweet functionality. It allows any account to pin into their own account. Pinterest will then pull that pin out of that account and shove it into random people’s feed… even if the pin originated from a now blocked account.

Still, blocking an account doesn’t do anything to block Pinterest’s crap relevance engine. Even if I block account by account, Pinterest’s engine insists on filling my feed with all manner of random garbage similar to what was blocked.

Following Accounts

You would also think that by following other Pinterest accounts, Pinterest would be more inclined to show us pins by those accounts whom we follow. Again, you’d be wrong. While Pinterest does show pins by followed accounts in the feed, it also intermixes in accounts not being followed. In fact, I’d say that Pinterest tends to show more account pins not being followed than those who are being followed. Sometimes that may have to do with when those followed accounts are active.

For example, if your followed accounts haven’t been active in the last hour or two, then Pinterest still insists on filling your feed with pins (a feature that is entirely unnecessary). If those I’m following haven’t pinned recently, then show me a blank page. It’s fine if the page has no pins. I’d rather see no pins in my feed than a bunch of random garbage.

Anyway, when pins by accounts you are following don’t appear in the feed, it could simply mean they’re not pinning. Instead, your feed is being cluttered by extraneous random garbage. The trouble is, it is truly garbage and not at all relevant. The weird thing is, there is so much more relevant content on Pinterest that the engine never finds and places into my feed. I have to use Pinterest’s search panel to go find it. It’s this random irrelevant garbage that makes Pinterest completely worthless as a platform.

You’d assume that Pinterest would prioritize followed account pins over random pins, but again you’d be wrong. Pinterest has no interest in trying to make their engine more relevant. They’re simply interested in promoting random accounts’ pins into feeds, even when those pins make absolutely no sense for that particular user (i.e., image of babies shown to grown single men).

The Pinterest Idea

The idea behind the Pinterest platform has merit. Too bad Pinterest’s implementation is such absolute shit. Images can be incredibly powerful, particularly so when that image is actually of interest to the viewer. On the other hand, images shown to people who have absolutely no interest in that subject matter is a wasted opportunity to show much more relevant content.

Pinterest wastes its opportunities every single time you refresh the page. Instead of feeding me actual images of interest, I get images of high heel shoes, of wedding dresses, of women in wedding dresses, of women’s hair, of babies, of smokey eye makeup, of tattooed body parts. I even get images of dog food bowls, dog collars and of dogs. I don’t own a dog. I no have interests in any of that. Yet, image after image after image is shown. It’s entirely frustrating dealing with Pinterest’s garbage.

But, that’s not the problem. Pinterest gives us NO TOOLS to actually wholesale remove these uninteresting photos from our feed. We have to deal with them one by one. We have to block accounts one by one. Even after going through all of the hoop jumping of blocking and reporting and hiding, photos of similar content STILL appear in the feed… day after day. Sometimes even the same pin I’ve reported or hidden STILL appears.

Just when I think I’ve got a handle on my feed, Pinterest re-ups and I get a whole new wave of garbage in my feed. With Pinterest, you simply cannot win that battle of spam photos. It’s a trash platform designed to be trashy. I’m amazed that it even still exists. I’m even more amazed that anyone finds it useful.

The Pinterest Dilemma

And here we come to the point that matters most. This is why Pinterest fails. The platform fails because Pinterest attempts to ‘guess’ what it thinks you want to see. Instead of actually asking you explicitly for interest categories, it attempts to learn what you like by the pins you click on. Unfortunately, it goes even deeper than that. It learns what you like by what those whom you follow click on… and those whom they follow click on. It feeds crap to you based on the interests and clicks of others, not what you specifically click on. It assumes that because somewhere down the line, someone you follow clicks on pictures of babies, you must also want to see pictures of babies or a bridal dress. This “sixth degrees of separation” assumption is entirely wrong for a relevance engine and needs to be removed. Of course, Pinterest also makes wrong assumptions simply by reviewing your activity.

When reviewing your personal activity, Pinterest’s difficulty is, like the tattooed fitness guys, its engine guesses wrong nearly every time. Instead of Pinterest seeing a bodybuilder in a fitness pose with a great physique, Pinterest sees the image as simplistically as a “person with a tattoo”. It then makes the entirely wrong assumption that “tattoo in image = interest in tattoos”.  It’s a simplistic, unsophisticated kindergarten assumption. It’s such a basic assumption, only a child could actually jump to that conclusion. Even then, only a child would jump to that conclusion if the parent already had tattoos and invited over a bodybuilder with tattoos. Only then might a child associate tattoo interest.

Having a relevance platform make the wrong assumption and jump to the most wrong conclusion is actually the worst of all possible outcomes for a relevance engine. It then leads your entire results astray and leads to frustration by what’s presented… thus making the platform worthless. It also means that once your “learning” machine learns this entirely wrong data, it’s doubly difficult to “unlearn” it. As I said, “Garbage In, Garbage Out.” In fact, Pinterest has no way to correct these completely incorrect assumptions its engine has made.

Pinterest could fix this by asking direct questions about pins to understand if the assumptions it has made about a specific pin is correct. If the assumption is incorrect, it can “unlearn” a learned assumption. Better, simply ask us what we want to see in our feed and exclude all else. Also, give us exclusion features. See a pin, click to exclude all similar pins from the feed. Even then, Pinterest still needs to get rid of its association algorithm where it associates “women in bridal outfits” or “babies in bassinets” or “doggie treats” with “men’s bodybuilding”… which is probably entirely attributed to its completely incorrect “six degrees of separation” relevance idea.

With all of that said, Pinterest does offer a mechanism to stop seeing pins “Inspired by”, but that’s a sledgehammer approach. Using that feature is all or nothing. It will stop the garbage, but it will also stop relevant pins. This feature is poorly designed and implemented. It’s the wrong approach for a relevance engine. Instead, as I said, as Pinterest users, we need exclusionary features that look at the image and exclude all like-kind images from the feed. Unfortunately, Pinterest just doesn’t get it!

[RANT OFF]

Since this is not only a rant and also doubles as a review of the Pinterest service, I rate Pinterest a solid 1.5 ★ out of 5. Pinterest, you seriously need to get your act together.

If you enjoy reading Randocity articles, please follow, like and share the article on your social media feeds. If you have had similar experiences with Pinterest, I’d like to hear your feedback via a comment below.

↩︎

 

 

%d bloggers like this: