Hacker News new | past | comments | ask | show | jobs | submit login
TikTok and the network effects of creativity (eugenewei.com)
136 points by MaximumMadness 11 days ago | hide | past | favorite | 66 comments

The most disruptive part of Tiktok is how it managed to dramatically reduce the "rich gets richer" effect of entertainment platforms. If you upload a video, TikTok will show it to other users even if it's your first. This allows them to assess it and progressively grow its audience. On YouTube and Instagram you'd have to rely on search traffic or external sources to build up your audience and get recommended (except for the new Reels / YouTube Shorts that are mimicking TikTok's UI).

I've done an experiment on the new year and created a video on a new account trying to catch people's attention (relevant to the new year, funny, with something unusual). I've done i on Instagram, TikTok, and YouTube. On TikTok it gained over half a million views, with 0 views on YouTube and Instagram.

Another way to frame this is around the Gini Coefficient of these platform. TikTok has a much lower inequality measure, which increases the incentives to produce content and hence the quality of content.

Youtube early days were similar to TikTok or any new platform, there are now just regular people doing creative things. As more and more videos are added, more businesses start to participate and more video depth/detail is added, then it gets harder to compete. It becomes less about just content value creation and more about content value extraction.

It is almost as if these platforms need a few different algorithm setups, the ones that are for more long term quality content, indie/business and just one that is raw where fresh content can be seen. They try this with "trending" type systems but due to market size they are usually heavily manipulated.

Right now featured platforms like TikTok are similar to like Twitter in that only the latest stuff matters. On Youtube it is more about quality, long term content for many creators. It is about right now but also you'll find amazing videos on history, art, gaming, development, markets, information and more. Youtube, Vimeo and others or sites that have been available for a while, are more about all types of creators so the real-time hype of new content isn't as successful for creatives, they start building more niche or long term bases that requires more work to produce. To compete the levels of production go up and up.

Shorter term real-time fresh platforms like TikTok are like the "new" algorithms on reddit/HN or the Twitter style freshness.

Longer term real-time fresh AND detailed deeper content videos are more like what you find on when looking for information on a search engine like Google or "popular" algorithms and more refined, more competitive and owned by larger players, but also more about information and answers.

Serious question - if we are being recommended anything to watch, shouldn't we have some sort of idea of what the algorithm does?

I would love to see advocates for openness in algorithms, as we have advocates for open source software. We have no idea what filters are applied by tiktok, youtube, etc. If I knew what they were, would I agree with them?

Questions that arise for me, are:

* What are the value judgements behind the algorithmic recommendations?

* Is it ok for corporations to entrain their users with specific content, if that it is not based on a neutral algorithm?

* There is surely interest on the part of corporations to promote or constrain certain ideas, that do not suit them. These would play out in the political, economic, legal domains. If the algorithm does not let you know, this would be a lie by omission, and therefore immoral IMO.


I don't think I have seen any discussion on this, albeit I think it is a hugely important issue.

To be fair, the discoverability on Youtube is shit and has been for years now (another topic).

As much as I dislike TikTok for being idiotic, that's actually an aspect to sort of like about it.

Of course it's still algorithmic spoon-feeding, but it seems they're inching closer to saner discovery tools. One can hope...

Are you sure this isn't more of the Medium model, where they can show you a bunch of new people all the time, but you'll never actually follow them or see them ever again (just another nebulous medium post)? So really the only person that profits in this model is the company hosting the content.

That happens with TikTok, but it’s a result of the app’s UX. (It seems intentional, but that’s conjecture.) Following a user whose video you like is easy, but the app defaults to the “For You“ page, which presents videos as picked by an algorithm. That algorithm doesn’t just iterate over all the videos out of a user’s followed artists, leading to the effect on creators mentioned. I may never see an artist’s new content if it doesn’t surface via the For You page.

Exactly. I think their algorithm explanation explicitly mentioned giving less weightage for follower counts and users following a particular creator. I'll not be surprised if the following feature is mostly for satisfying creator egos (thus helping to keep them on the platform) and not for the algorithm itself.

One excellent feature of the algorithm is how it ties user feedback to the type of content shown and not to the creator of the content, thus giving better recommendations.

When I visit TikTok without logging in I see a side bar recommending Will Smith, Gordon Ramsey, Kevin Hart, Selena Gomez, and SnoopDogg.

They may surface other stuff better than other platforms, but they definitely have the same "rich gets richer" stuff in the default landing page too.

Well, I am probably not getting the argument that you are trying to make. When you need to surface content without any idea what the person might like then using aggregates of popular content for a given culture is a good low risk default

I think he was expecting popular content from unknown/non mainstream artists instead of popular content from popular artists that can be found anywhere.

This always sounded pretty unequal to me.. has it stopped?


https://en.wikipedia.org/wiki/List_of_most-followed_TikTok_a... does have less traditional celebrities compared to the lists for other social media platforms

This is an issue I've always hated about Steemit.com, too.

There were a few writing mediocre at best content making up to $1,500 a day/post at some point, just because they managed to gain a lot of followers initially, while everyone else had to deal with writing dozens of posts and making only a few dollars a month, if that.

Social platforms should take this "inequality" issue a lot more seriously. It's in their long-term benefit to have thousands, millions of people who "make it big" on their platform, not just a handful that make it "really big."

I watch TikTok and watch their videos because my audience is just me. And because they are just me. I don't want to watch anyone else because I believe it's not worth watching them. But they are my audience. That's all I want. And they are my audience. And I feel like I should be making sure that they feel that way for me.

And they are making damn sure that it is worth watching.

When I think of this "I believe this thing is fucking worth watching" line I think of the way things used to be.


Do you mean the model? What about it?

I think they mean the text of your comment. I’m unable to parse it.


I am not an GPT-3 robot and I do not even having access to API or model while I am on the wait listing. I do not speak of native ENglish and maybe no great writer. Please excuse this deficiency.

Every video I have seen that managed to escape TikTok's confines and infiltrate my consciousness could be inserted into an episode of 'America's Funniest home Videos' and leave nobody the wiser.

And I'm sure they're frothing at the thought of all the content they can copy-paste into their show.

they have an account @afvofficial

I think the "each new member can enhance creativity of existing members" vibe can exist generally among "scenes", and scenes definitely can use technology that directly taps into this.

I like the "scenius" term for this from Brian Eno: https://youtu.be/0qATeJcL1XQ (54:15 in that talk)

TikTok describes itself as: "The most fun, honest, true and honest place on the web you will ever get." That's how TikTok feels. The thing about TikTok is that, like with Facebook itself, you know every detail right from the beginning. Your friends and everyone who you have ever interacted with on Facebook are your peers. But TikTok is more powerful than Facebook when it comes to the social fabric. TikTok is like a community hub with social networks where everyone can be themselves. There are no limits when it comes to who can do what and how they can interact with each other.

But there are limits. Nudity for one. Sexual content. Nevermind the more serious beheadings or jihad videos.

I don’t know why people are underestimating Instagram Reels. Instagram may be becoming more of an entertainment app like TikTok but it is still a social network. All content creators care about is engagement, and having the ability to push your short video into your Stories and your follower’s main feed increases your chances of it being seen instead of relying on an algorithm.

Instagram Reels might have the same concept but a.) different demographic, with TikTok gaining the lion's share of the younger generation and b.) recommendation algorithm, where TikTok explicitly pushes you newer content from barely known creators, while Instagram relies on the Facebook way of likes, shares, comments, hence giving you the same content from the same celebrities over and over.

You do rely on an algorithms on the story feed and main feed.

Reels are kinda misplaced. They are so short but require so much navigation. Why not just post a short video. Ig live videos are more important and need the first class treatment (a seperate search, a ig live discovery page)

This comment is a slightly tangential shameless plug, but I just want to point out that I loved the structure of this essay. There were a lot of great little ideas that I don't think would have seen the light of day in a more traditional sort of blog post.

> All the points I wanted to cover seem hyperlinked in a sprawling loose tangle. This could easily have been several standalone posts. I've been stuck on how to structure it.

> This piece is long, but if you get bored in any one section, you can just scroll on the next one; they're separated by horizontal rules for easy visual scanning. You can also read them out of order. There are lots of cross-references, though, so if you skip some of the segments, others may not make complete sense. However, it’s ultimately not a big deal.

I've long wondered how many essays don't get published because the author struggles to generate a "through line." Sometimes organizing ideas is harder than coming up with them. This is certainly a problem I struggle with.

I've been working on software[1] that encourages you to publish ambitious online media even if it's a bit disjointed. Currently, only me and my friends and family are using it because it is very rough around the edges, but it is good enough that I personally use it every single day.

If this sounds interesting to any HN comment readers I'd love to give you a beta code or a live demo to hear your thoughts. Send me an email at jon@edifice.pub

[1] https://edifice.pub

What kind of format would it be published in?

This is pretty interesting as it is a problem I struggle with too.

Your landing page there reminds me of something else I’m a fan of, e.g. Building a Second Brain through notes and web clips as discussed here: https://tanners.blog/diy-second-brain/

But these are more focused solely on the background processes of organizing information.

I’m curious about how your system would encourage publishing, “even if it’s a bit disjointed”, as that seems kind of the logical next step. (And definitely a challenge for me too!)

I'm a fan of Building a Second Brain too, and I've taken a lot of inspiration from it. In fact, the precise reason I'm working on this project is because "organizing information as a background process" feels a bit too much like navel gazing to me and things are a lot more fun when the process includes interacting and sharing with other people.

The particular way I'm attempting to encourage publishing is an editor that lets you lay out your stuff as a directed graph instead of as a linked list. The reader navigates the graph by clicking on hyperlinks that take them in tangential directions that may or may not converge with the main thread.

The inspiration for this includes those really good conversations you have with your friends where you go completely off topic but in an interesting way, and also the experience of going down a wikipedia or tvtropes rabbit hole that consumes hours of your time.

It's a tough problem because your eyes and ears are only able to process information serially, but concept-space is a complex multi-dimensional snaggle. How do we bridge that gap? The conventional answer is "good writing" but that's really tough to do. Eugene Wei didn't feel up to the task with the ideas in the OP, for example. I'm hoping that by giving people more room to play with how they structure their thoughts and ideas "writing skill" will be less of a constraint on human communication. I definitely don't think I've hit an optimal solution yet but it is a very fun problem to work on and talk about!

It sounds like a tremendous amount of creativity, but limited in scope to making videos? It seems kind of narrowly focused on entertainment.

Sorta but in a lot of TikTok niches the endgame isn't really a video, that's just how it's presented. All the hobbyist communities make videos that show off their work and inspire others in their bubble. Like sure the video is "entertainment" in that it holds your attention but that's pretty much the extent of it.

Like hobby TikTok is genuinely just a bunch of nerds being genuinely excited to show you something they're passionate about or something they made. It's so god damn refreshing. Like the early days of Tumblr.

...or anything else that you can fit into video, such as news and politics or education and editorial.

Eugene Wei on a16z podcast https://a16z.com/author/eugene-wei/

The author is very incorrect.

> By network effects of creativity, I mean that every additional user on TikTok makes every other user more creative.

That is false.

TikTok isn't a creativity amplification network, it's a mimic network. The extreme majority of humans are mimics, they essentially never create or do anything creative or original. They are incapable of that (cue the outrage at such a statement, even though it's true). They play follow the leader across a lifetime. TikTok, like most social networks, represents that accurately. What TikTok does not represent, is a burst of individual creativity that is widespread.

It's a creativity distribution channel. The 0.01% that are originators distribute creativity to the drone mimics and they copy and share it.

That's exactly what the dance copying represents for example. There is no great creativity explosion going on there, quite the opposite. As with YouTube or any other distribution system, an exceptionally tiny percentage of people are originators, actually creative, the rest mimic and pander and try to scam their way to some views by copying or ripping off originators (you see this repeating trend represented in everything, eg content farms).

> TikTok isn't a creativity amplification network, it's a mimic network.

Replication with some source of variation (such as even just “people aren't perfect mimics”, but “some people apply some minor modicum of creativity” enhances this) plus selective pressure (such as interesting novelty getting rewarded) is sufficient for it to function as as Darwinian creativity amplification system by way of being principally a mimic network.

"Everything is a remix"

TikTok being a "mimic network" does not show it's not creative.

Creativity is primarily a product of creation, rather than one of originality and TikTok is, in fact, inspiring to create.

Recognising creativity is also creative. Art is in the eye of the beholder (not necessarily the creator).

Something can be authored that was unintentionally interesting, but amplified by the creative recognition of some weird facet. For example cat videos - the cat isn’t trying to get likes! Edit: or a security camera video where there is zero intent to create, and the art is in the recognition of the clip by everybody.

I suspect that out of a zillion mimic events you will get a small number of truly creative bursts, yes. I think it's very rare though and wouldn't qualify as being widespread creativity or origination. Also those bursts of new/original content coming out of initial copying may primarily come from burgeoning originators that are just being born so to speak, rather than from the mimic group. I doubt the process overall results in an increase in originators.

edit: to the repeat downvoters instantly hitting every one of my comments, those that don't like the fact that I'm pointing out that humans are 99%+ mimics, I'd encourage you to add to the discussion and dissent from what I'm saying. I'd enjoy reading that counter. Everyone isn't a butterfly just waiting to be unleashed into the next da Vinci, that's a fantasy. It makes perfect sense that the majority operate as distributors, mimics, for things that the 0.001% come up with that work effectively. It would be an enormous biological waste of energy for everyone to be so creative, the mimic and distribute what works approach is logical. Humans do it with everything, including learning / copying skills, behaviors, systems, almost anything you can name. For example, there are always a very small number of teachers (as a share of the population) distributing knowledge/skills, and most teachers are also mimics, but they're custodian mimics that use various bullhorns to (ideally) spread what works faster. Teachers are rarely originators, the knowledge is passed down a distribution chain by mimics that serve various functions along the way. That's how a lot of systems in human societies work (politics and religion all work that way).

This reads to me like one of those 90's articles over-simplifying and underestimating the internet. Not that what you're saying is completely wrong, and trust me I'm not saying TikTok == the Internet, but the belittling tone is off-putting.

I'd have to put a lot of effort into walking around the inevitable tone that is given off by calling the majority of people mimics. I know how that would have to come across. I don't think it's worth taking that long of a stroll to reach that outcome, the people that are going to disagree and instantly downvote due to outrage are never going to agree with the mimic premise no matter what I say or how I tone it.

I don't view the premise as bad at all, or negative. It makes sense, it's a very reasonable biological system of replicating/copying what works and passing it along; it's very energy efficient for a species, and historically we've had to conserve energy, our evolution would strongly favor a mimic what works system. As a concept it also doesn't elevate us above other biological systems that operate on this planet, we're not that special; whereas to pretend that everyone can be da Vinci is to falsify what humans are, to pretend every person is a creativity giant in waiting if only they got the right encouragement. It's just more of everyone gets a trophy culture in action.

I think you're wrong, but I don't think this should be downvoted.

I think it helps to compare "copy and remix" with "does nothing". Compare "watched 50 difference dances" with "saw 0 dances". When you see it in that light, it would be difficult to argue that some people don't get inspired by seeing new things.

Now, I might argue that the level of creativity added from each additional user is not linear. That the creativity added is vacuous, pointless, 99% mimicry, and doesn't actually move the conversation forward...

But there are those 1% of users who are able to participate in the global conversation and say something new, who otherwise would not be able to participate, that I think it's at least worth acknolwedging.

A few questions for you, because I think you make a good point about it being a mimic network (where each person puts a slight spin on it).

1. Have you heard of the belief of "Human Design"? That belief / theory / spiritual guide has a similar distribution to what you're talking about. I'm just curious if your thinking was independent to that or if you've ever heard of it. (https://www.jovianarchive.com/Human_Design/Types)

2. How do you define originality? Is it something where "you know it when you see it", or is it something else?

3. Thinking of probabilities, wouldn't it also make sense that the default for people is to be un-original? We only have so many elements and so many places, I have to imagine that people being 'incapable of [doing anything creative or original]' is less a value statement and more just a logical progression of probability.

4. I've also come to a feeling that a person doing something that they've seen before, but do not see right at that very minute, is 'non-internet brain thinking'. We didn't have visual records so prevalent until just recently in humanity's lifetime. Repetition of behavior (lately sometimes called 'holding space for X') is a useful function for social networks, giving validity to someone's creativity.

Again, not trying to invalidate anything you're saying (and I noted your 'cue the outrage' comment, hence why I'm over specifying this too)

Super interesting take on the article, I've been wondering similar things for a while.

"good artists copy; great artists steal"

Even someone as great as Picasso would not have been able to create the masterwork he left us without learning and getting inspirations (aka, steal) from other artists before his time.

If you ever think someone is creating good original art (because shitty original art is easy to create), chances are you are just not familiar with the work they stole from.

Quite the elitist view! When I watch a few TikTok videos I quickly form an opinion whether something is creative or mere copying, I know it when I see it. Just writing the whole thing off seems insane, statistically speaking.

TikTok is powered by “yes, and”, which is a perfectly fine form of creativity.

In its basic form, creativity is an act of creation. Whether or not these creations are at the Picasso/Mozart/Proust level of high art, creativity is technically operating in this network. Measuring creativity in the way where more of x produces more of y (more users=more creativity) is of course, not a real analysis. More people produce more content which may or may not be creative in nature (it can be a product review for instance).

Good point. Even here in HN is very common from n-th YC me too company, to the "check out my innovative new JS framework/library/tool".

Maybe all creativity is just mimics who repeat things and add a little bit of variation.

I think the book Steal Like an Artist addresses this.

Or maybe it isnt, and the great breakthroughs are the result of fantastic imaginative and courageous minds.

I dont think this is true. Creativity happens when you combine/blend prior knowledge. We don't magically create new information without prior meaning.

TikTok acts likes this and the remix is like a blending/combining prior creativity that leads to continuous new creativity.

This very easily refuted, then no creativity is possible because who was going to create the "first knowledge"

Your refutations is even more easily refuted. First knowledge is created by observing the world around us. Mimickry of the natural world is all around us, you even see it in animals.

So you could create the knowledge without imitating others. Try again. The second time maybe will be better. That without entering in the discredited philosophical position of empiricism for epistemology.

Fantastic and courageous minds working from the void?

Or fantastic and courageous minds responding to some cultural or historical material?

(Which would be much in the way a TikToker responds to a meme while adding some of their own spice).

Let’s not get bogged down by the fact that one domain might be perceived as more valuable/respected and the other not so.

What have philosophers ever done post-Socrates except come across the ecosystem of historical philosophers and riff off of that?

Yes, fantastic and courageous minds. The kind of person we all due our civilization: Newton, Hooke, Darwin,Gauss, Euler,Watts.

I think there is some truth in that those characters were more creative than many other, but they did not create in a void. Your first example, Newton, said "If I have seen further it is by standing on the shoulders of Giants." Who am I to contradict Newton? :)

Arent you the same kid saying in other comment that there is not creativity, just imitation? Take a position and stick to it. BTW you are misinterpreting Newton (on purpose or not). He didnt say he copy them, he was humbly acknowledging there have been giants in science before him.

Well, I think the insight in the article here is that a network (including such “giants” that came before) can boost the creativity of those who come after, even including a new giant such as Newton.

Think about any modern scientist. Has their creativity been boosted by the existence of all the other scientists?

If I’m a physicist, the existence of Newton boosted my creativity because now I can apply calculus to problems. If I’m a biologist, Im leaning on Darwin and all the insights of statisticians and the people who have delineated the methods and findings around which I organize my science.

People have previously innovated the concepts, methods, and foundational understandings, and finding new insight can be a matter of combining these in novel ways.

I don’t think this is completely different than what the author is describing in TikTok. Both the format and ecosystem of tools that the service creates, and the social network of users create an environment where it’s easier to go out and make something.

A similar thing happens in science.

Sounds like you're talking about the difference between a sourcerer and a wizard in Discworld.

In TikTok, it is easy to dismiss the whole project because, well, it's not a huge deal. It's fun. It's easy to take for granted that the entire thing is the work of a single person who is basically a public relations and marketing specialist named TiKTok. This is not a joke. This is a reality that requires every minute of the day to be spent creating an avatar for their fan base of 6.5M.

But TiKTok has only just launched that avatar. TiKTok has created another one. They are still creating the first and are hoping to add another.

And as a result, TikTok has become this strange thing called "the Internet's Most Dangerous Video." For some reason, even people who are a little skeptical have found the world and become so accustomed to the idea of TikTok and their little thing that they can't imagine how this project can kill their brains for some of the crazy shit they're actually watching on it.

As a corollary to that, genuine creativity requires a lot more time and effort to appreciate than a 60 second vertical video.

Most of social media is just shallow attention grabbing time sinks. IOW, attention hacks.

I generally agree with that. It's one of the things people loved about TikTok, they got to feel like they were doing something creative or original, producing creative content, without having to invest years into learning how to actually dance. Copy some basic arm motions and movements, repeat it a lot until you memorize it, then record it and publish it. It requires very little thought, and the young crowd that is mostly doing it has a huge surplus of physical energy to put into copying & playing back mentally empty physical movement.

> The extreme majority of humans are mimics, they essentially never create or do anything creative or original. They are incapable of that (cue the outrage at such a statement, even though it's true).

Of course people are going to be outraged over the statement. You're playing a semantic game where you came up with your own definition of creativity that excludes what most people do, and then not actually defining it or justifying it. If you used a more common definition of creative you'd have no argument to make.

Novelty isn't the only thing that makes something creative, a string of randomly generated numbers is always novel but wouldn't be considered creative. Creativity requires novelty conveyed through recognizable patterns, which means some amount of mimicry has to be involved.

Every dictionary definition of the word I’m looking at right now includes the words “original”, “originality”, or “not imitated.”

I think the word “prolific” applies to what TikTok facilitates, much more than the word “creative.” And personally I find the amount of word for word and beat for beat mimicry on the app maddening.

> Every dictionary definition of the word I’m looking at right now includes the words “original”, “originality”, or “not imitated.”

Citing dictionary definitions just shifts your argument from one word whose meaning you didn't define to another word whose meaning you haven't defined. Is the Mona Lisa original? It's not the first oil painting, nor the first portrait, nor even the first painting of an Italian noblewoman made in the renaissance style. Yet it's commonly considered to be creative despite being imitative and unoriginal in all those previously mentioned aspects. What definition of original can you then make that simultaneously includes the Mona Lisa, is mutually exclusive with any imitation, and excludes the "majority of humans".

Applications are open for YC Summer 2021

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact