[HN Gopher] Watch this guy work and you'll finally understand th...
___________________________________________________________________
Watch this guy work and you'll finally understand the TikTok era
Author : mooreds
Score : 70 points
Date : 2023-10-28 15:09 UTC (7 hours ago)
(HTM) web link (www.wired.com)
(TXT) w3m dump (www.wired.com)
| derekja wrote:
| https://web.archive.org/web/20231021002437/https://www.wired...
| withinboredom wrote:
| Thanks. I apparently "ran out of free articles" despite not
| going to Wired.com in years.
| Liquix wrote:
| Does this type of practice really convert casual readers to
| paying customers? It seems like it would just drive traffic
| away from the site.
| withinboredom wrote:
| Maybe. I refuse to pay for news though, on principle. I
| grew up with my dad working for the newspaper (in sales)
| and as far as I'm concerned, they make plenty of money
| through those channels and they don't need my money. If
| they need my money, then they're doing something wrong, got
| greedy, are "too big for their britches" (so to speak), or
| some combination of the above.
|
| Other people feel differently, and that's fine.
| earthscienceman wrote:
| I understand this sentiment but it's extremely short
| sighted. Paying for news is about incentives. Much like
| Google having poor incentives to provide good unbiased
| search functionality based on the fact that their revenue
| comes from advertising, news work poorly aligns with
| advertising incentives also.
|
| Whether you contribute to that or not is your choice.
| spondylosaurus wrote:
| One reason I subscribe to the New Yorker (beyond enjoying
| their content) is that they have legendarily high
| standards around fact-checking, which isn't something you
| can maintain unless you're paying people to do it. So I'd
| like to help them continue paying the people who do it.
| bobsmooth wrote:
| HNers will pay $10 a month for a google replacement but
| won't support the foundation of democracy for a few
| bucks.
| hotpotamus wrote:
| My mother has worked in local journalism for half a
| century - mostly radio, but some newspaper and the very
| occasional TV appearance. If they're making a lot of
| money, she sure isn't seeing it. Maybe whatever private
| equity company that owns the station this year is doing
| well, but they're not known for sharing with the other
| boys and girls so to speak. It's a shame because from
| what I can tell, she was a very good reporter but the
| role is reduced to a shadow of what it was when I was a
| kid. Obviously they're now talking about replacing her
| with AI, and she's about ready for a severance check I
| think.
| withinboredom wrote:
| I'd feel differently if and when journalism is as
| transparent about where the money goes (show the budget,
| even if just %'s). If I knew that my $10 subscription was
| actually going to journalism ... I'd pay $10.
|
| But it's not transparent, and those sales guys/gals are
| making bank and who knows where the money is actually
| going. Clearly, not the journalists.
| ssalka wrote:
| all it's taught me to do is develop muscle memory for
| looking for an archive link in the comments or checking
| wayback machine
| IAmGraydon wrote:
| It makes no sense to me when it comes to actual news
| because I can usually get the same info from 10 other sites
| for free. For unique content, I could see it working. My
| problem is when these sites charge to see the articles and
| still run ads right in the middle of the editorial. You get
| to do one or the other, not both.
| underseacables wrote:
| I was a subscriber to wired magazine for over 10 years,
| but they kept reducing the articles and increasing the
| advertisements, to the point where 80% of the magazine
| was just advertisements.
| rqtwteye wrote:
| I think it drives people away. A micropayment system with a
| few cents per article would be way better in my opinion.
| fiddlerwoaroof wrote:
| Yeah, I really liked Blendle for this but they recently
| shut down US operations for some reason.
| selfhoster11 wrote:
| I think the reason being these were more like
| "minipayments" or "centipayments", and not micropayments.
| When I used them, there was a real mental friction for
| deciding whether a particular article will be worth the
| $0.07-0.21 that it was priced at.
|
| A far better model is either do actual _micro_ payments,
| meaning sub-1 cent so you don 't feel ripped off if an
| article was not what you were after, or to let you pay
| what you want after reading the article.
|
| The economics of making this work is left as an exercise
| to the reader.
| fiddlerwoaroof wrote:
| I think what I'd like is a service that's about $14/mo
| with all you can read news. Then something like $7 of
| your subscription fee is divided across sources based on
| the amount you read them, tidal-style
| underseacables wrote:
| I think it reinforces the resistance mindset. Humans don't
| like barriers to something they want, and they get
| satisfaction from getting around the barrier. The more
| often they do it, the more reinforcing it becomes
| teaearlgraycold wrote:
| VPN?
| IAmGraydon wrote:
| If you're viewing on a mobile device, turn on reader mode and
| you can get the entire article. It doesn't work for all
| paywalls, but appears to work for this one.
| Kiro wrote:
| I "ran out" when switching back to the tab in the middle of
| the article.
| paulryanrogers wrote:
| It all seems precarious and a bit shady
| musicale wrote:
| Gaming the algorithm can probably work for a while, at least
| until the algorithm or platform changes.
|
| Consider how successful a number of shady SEO practices were
| until Google shut them down. Then Google stopped caring and
| simply pushed everything below the fold of Google promotions
| and advertisements.
| PartiallyTyped wrote:
| Here's my question then, let us assume I have some algo to
| churn out content [1]; is there any reason I shouldn't do
| some genetic programming and optimize the process based on
| income to chase the algo?
|
| [1] https://www.youtube.com/watch?v=IcF9H6ZzWu8
| adolph wrote:
| That video is kind of hilarious in that he kept rewriting
| things he wasn't happy with instead of focusing on the
| pipeline. That is the opposite of just letting genetic
| programming perform the fitting function.
| PartiallyTyped wrote:
| I am saying, if I can create a pipeline - which for
| shorts and simple stuff is fairly trivial - why not let
| GP deal with algo chasing?
| kodapoda wrote:
| I used to feel more triggered by these practices, until I
| mentally decoupled art and creativity from whatever this is
| ("content" I guess). This is just its own world of gaming
| "engagement" and getting paid for churning out "content". The
| incentives are aligned and even encourage the practices
| described in the article. In that environment it's an expected
| course of action, not necessarily shady.
| TeMPOraL wrote:
| > _In that environment it 's an expected course of action,
| not necessarily shady._
|
| It isn't shady to the participants, in the same way that the
| activities of an organized fraud group aren't particularly
| shady _to the members of that fraud group_. They are to their
| victims, though.
| kodapoda wrote:
| The victims in this environment are people who will watch a
| 1-minute long workout video featuring background music from
| X "three folders of content" vs Y "three folders of
| content". The algorithms have been manipulated, sure, but
| the actual negative impact is close to zero.
| spacecadet wrote:
| Castles made of sand.
| esafak wrote:
| How to hustle hard in return for little money and no career
| path.
| Towaway69 wrote:
| The moment you realise the career ladder is a hamster wheel.
| IAmGraydon wrote:
| There's massive survivorship bias going on in this space
| right now. Social media is basically a survivorship bias
| machine, after all. You never see the many thousands who try
| this and completely flop.
| rchaud wrote:
| That's every market with an inelastic supply of labour. P
| Diddy could reboot Yo! MTV Raps with a million-dollar cash
| prize, but that's not going to uncover another Biggie.
| dehrmann wrote:
| I'm used to elasticity in terms of something like oil
| where people will pay for it, whatever the cost. I guess
| creative endeavors are the same, just on the other side.
| There will always be another artist willing to work for
| nothing in hopes of getting discovered.
| esafak wrote:
| The economic model for that is a winner-takes-all market,
| which media is a textbook case of. Social media, on the
| other hand, is a little heavier tailed, but still nothing
| like engineering.
| anigbrowl wrote:
| Snobbery is so unbecoming.
| kodapoda wrote:
| This is just another way of saying everyone who doesn't
| like your "content" is a "hater".
| cookie_monsta wrote:
| Yes, let this be a warning to trap metal artists everywhere:
| the quarter million dollar record deals are a smokescreen.
| Solar panel sales are where the real action is.
| rchaud wrote:
| That's the headline number record labels and A&R men use to
| dupe young musicians with no familiarity with corporate
| America.
|
| That cool quarter million is an advance, and will end up
| being used to offset the logistical and marketing costs of
| recording an album and going on tour, which are both
| organized by the label. So really, the artist is just
| temporarily holding on to money that will be paid back to the
| label.
| Eisenstein wrote:
| Do you know this for a fact or are you interpolating based
| on legacy music media tactics?
| cookie_monsta wrote:
| a) Sure, but that's the game the music industry has been
| running ever since they concocted this scam. It has nothing
| to do with how an artist got on their radar in the first
| place
|
| b) Still sounds better than selling solar panels, to me at
| least
| rufus_foreman wrote:
| Fall in the sea
| lotsofpulp wrote:
| > and that she uploads three folders' worth of content
|
| Sounds like a rigorous metric.
| lionkor wrote:
| This is when you know the system you built has reached peak
| ubiquity
| dylan604 wrote:
| what exactly is a folder's worth of content so that I'm
| impressed by it being 3 of them?
| akira2501 wrote:
| > It started with me ditching journalism to focus on churning out
| daily TikToks in which I'd offer tips about storytelling.
| Magana's talent management startup, 25/7 Media, would ensure
| eyeballs for this content by enlisting its 60-plus clients to
| drive traffic my way. Once I'd built a decent fan base, 25/7
| would produce a weekly podcast featuring my candid conversations
| with up-and-coming digital creators.
|
| So.. it's a social media ponzi scheme?
| DistractionRect wrote:
| I wouldn't necessarily call it that. IMO, it's mostly about
| juicing discovery. There's more content than any one person
| could ever consume, and a limited attention economy. Basically,
| they find what they believe is good content, and bring it to
| the foreground. If there's no audience for it, it won't build a
| following and just fall on it's face. If it takes off, they
| have to capitalize on the momentum and produce content to stay
| in the spotlight.
| haltist wrote:
| Most digital content converges onto optimized spam because
| that's what grabs the most attention on any given platform.
| Spam is basically an algorithmic exploit which is the main
| reason for why delegating decisions about what should be
| worthy of attention to commercial algorithms designed to
| optimize advertising revenue is not a great idea. The overall
| effect is detrimental because it trains people to stop caring
| about what is fact vs social fiction (e.g. conspiracy
| theories).
|
| Lots of folks have written about the detrimental effects of
| revenue optimizing statistical algorithms and their use in
| social media. For anyone who wants to learn about this I
| recommend Jaron Lanier's book. He does a good job of
| explaining the reasons for not delegating one's choices to
| social media algorithms.
|
| Edit: the above comment is AI approved. Pasting the response
| from claude.ai for posterity.
|
| You raise some thoughtful points about the problems with
| algorithmic content optimization in social media and digital
| platforms. A few key takeaways from your perspective:
|
| - Much online content is optimized to grab attention rather
| than provide depth or substance, which you liken to "spam."
| This is driven by algorithms designed to maximize engagement
| and ad revenue.
|
| - Over-reliance on these algorithms can train people to care
| less about facts vs fiction/conspiracies. It promotes shallow
| thinking and engagement over meaningful discourse.
|
| - Many experts, like Jaron Lanier, have warned about giving
| too much influence over to revenue-driven algorithms in
| determining what people see and engage with online.
|
| - There are real risks and downsides to letting algorithms
| choose our information diets. It can promote conspiracy
| theories, erode critical thinking, and undermine a shared
| sense of truth/reality.
|
| - You advise more conscious thought about how we interact
| with algorithmically-optimized content, and highlighting
| writers like Lanier who provide deeper analysis of these
| issues.
|
| I think you summarize the challenges very well. It's an
| important discussion as more of our information and
| interactions move onto digitally-mediated platforms. There
| are no easy answers, but being aware of the risks and
| downsides is a good first step.
| cookie_monsta wrote:
| > the above comment is AI approved
|
| I hope I live to see the day when this actually means
| something
| haltist wrote:
| The simplest examples are social media algorithms. The
| content shown to users is vetted and approved by the ML
| models to optimize engagement and advertising revenue.
| cookie_monsta wrote:
| Great, but using it to fact check your own comments on
| HN? When will that become viable?
| haltist wrote:
| Every AI service has an API so it would be easy to write
| a driver for a headless comment bot that checks comments
| with AI and then replies or updates a comment based on
| the API results.
|
| I think bots are against the rules on HN but the
| implementation would be very easy if the bot didn't get
| banned.
| akira2501 wrote:
| That's why I highlighted that quote. I see it as a suggestion
| that he can take someone with negative presence currently and
| with no drive for it into social media fame by merely
| applying this formula, and that part of this formula would be
| repeating the same step for new recruits.
|
| Perhaps it's just the way his marketing bluster read into the
| article, but it suggests he's not concerned with the market
| forces you're invoking, and so I immediately saw a large
| pyramidal structure when I read that particular line.
| gtvwill wrote:
| By formula you mean bot farm...
| simonw wrote:
| That feels to me more like a network effect than a pyramid
| scheme: help your talent out by having them collaborate
| with each other.
| escapedmoose wrote:
| That seemed resolved by the end of the article, by the
| admission that this guy sometimes has 9 bad ideas in a row
| and the 10th is a winner. The highlighted example was one
| of the bad ideas, but it illustrates generally how the
| promotion process works
| kbenson wrote:
| Using a network or community of people to promote each
| other is not a ponzi scheme. A ponzi scheme would be if he
| took profits from those new people to somehow pay the old
| people. This is actually the opposite of that, where he's
| getting time and effort from established people to boost
| new people.
|
| It's functionally no different than having a friend in an
| industry that gives you a shout out when you start a new
| venture, except you're opting into a group of people all
| doing this to help each other. It's nothing new, and
| nothing problematic IMO, and it's been going on for
| hundreds of years at least, and possibly thousands.
| adolph wrote:
| > a limited attention economy
|
| It would be ironic if the content they generate is shortening
| attention spans, thus making the overall economy smaller.
| TeMPOraL wrote:
| Except it's all effectively a self-referential loop of fake,
| empty content that pretends to be of high quality, but exists
| and is invested in only to promote other such fake, empty
| content. They pump energy into the loop to grow it - which
| incidentally displaces all topically adjacent but authentic
| and useful works - and grow it until it becomes self-
| sustaining. The _only reason_ this is done, the only way this
| actually benefits the people involved in this scheme, the
| only reason they do this, is because companies pay for
| getting themselves attached to the content in the loop. _It
| 's all just a more concealed form of advertising._ Any value
| the content in question has is entirely incidental, and
| quickly optimized away.
| IAmGraydon wrote:
| The sentence about driving traffic with 60-plus clients does
| have a Ponzi-like quality to it. It's only sustainable if he
| keeps adding clients at a faster and faster rate. At least
| that's how it appears to me.
| bobsmooth wrote:
| The point is to get the ball rolling until organic growth is
| achieved.
| lionkor wrote:
| Yes, but thats neither illegal nor are people being exploited
| (not directly, anyways) so this kind of ponzi scheme may be as
| close to legitimate as it can get(?)
| Bukhmanizer wrote:
| IMO this is a weak criticism. It's pretty common for eg.
| podcasts within the same network to promote a new podcast by
| pointing their listeners towards it. Is it annoying? Yes. Is it
| a Ponzi scheme? No.
| infamouscow wrote:
| I too can make brain dead straw-man comments that sound pithy
| in a pathetic attempt to derail the topic at hand.
| 0xDEAFBEAD wrote:
| I think it would only be a Ponzi if social media shout-outs
| were zero-sum. But they're not. I can tell my audience to check
| out your stuff, you tell your audience to check out my stuff,
| and there's the potential for mutual benefit. Maybe some
| disengaged people in my audience engage much more with your
| content. Or maybe our followers just spend more time on Tiktok
| overall, or more time on our Tiktoks as opposed to other
| peoples' Tiktoks.
|
| Another point is the rich get richer on social media. Having
| followers is how you gain followers. Social proof is important.
| People like to talk about what everyone else is talking about.
| Once a new client reaches critical mass, they're going to
| passively gain followers due to this effect (e.g. creators with
| more followers are more likely to be recommended by the
| platform -- at least I would assume this is generally the
| case). So adding in a new client and gaining followers for them
| is an investment in the long term.
|
| I think YC companies buying services from each other could be a
| good analogy actually.
| markhahn wrote:
| Bleah
|
| I guess I'm old because I don't even find it zoologically
| interesting.
| djmips wrote:
| That doesn't mean you're old, it means you're burned out.
| groestl wrote:
| > [...] articles never go viral
|
| I mean, it's on the front page of Hacker News now, does that
| count?
| simonw wrote:
| A Hacker News homepage link drives 10,000 to 20,000 impressions
| in my experience, which is peanuts compared to a viral TikTok.
| adolph wrote:
| Wait, people are making tiktoks about your best TILs or
| something?
| simonw wrote:
| Sadly not - but TikTok stats are visible so you can see how
| popular things get and it's way more than tens of thousands
| of views for anything successful.
| kbenson wrote:
| You just made me think about what it would be like for
| people to take and summarize the interesting submissions
| here on tiktok, and more importantly the interesting chunks
| of comments, and I definitely would watch that if well
| done.
|
| It would take a lot of skill to do well though, and I think
| that's the secret sauce, as I would read it just as easily
| as watch it if it existed in that "done well" state.
| jvalencia wrote:
| If you drove 10-12k views, the algos on TikTok would notice
| and it probably boosts well beyond the initial bump.
| Solvency wrote:
| Maybe ironic but, I don't watch TikTok or Instagram so I blame
| ChatGPT for killing my attention span: I used to enjoy reading
| Wired articles (and others) a decade ago.
|
| Now, I can't stand all the irrelevant prose about the authors'
| personal experiences, how they met the subject in a cafe, and so
| many other extraneous details injected in for padding and
| storytelling. Hunter S Thomson did this decades ago.
|
| If this is about how some TikTok manager games the system, a
| simple bullet list would be great...and a worthy length given its
| subject matter.
| KineticLensman wrote:
| Hunter S Thompson's interjections _were_ the story. Remember,
| he wasn't the fly on the wall, but the fly in the ointment.
| mihaich wrote:
| That's an interesting observation and it applies to me too. I
| just asked ChatGPT to summarize it.
| RoyalHenOil wrote:
| I have never in my life had any patience for those meandering
| sorts of articles. And yet I eat up novels full of detail, and
| I have an endless attention span for reading comment sections
| like this one.
|
| At the end of the day, it comes down to what I'm looking for:
| information or entertainment.
|
| When I read a new article (or a recipe, etc.), I do not want to
| be entertained: I am after specific information.
|
| When I read a novel, though, I absolutely do expect to be
| entertained, so I enjoy all the fluff.
|
| When I used to read print magazines and papers, they fell into
| the "entertainment" category. But when I read on a screen,
| that's not what I'm after; reading on a screen is an inherently
| uncomfortable process and better suited to bite-sized chunks.
| rchaud wrote:
| Perhaps you could get ChatGPT to read the article, and then
| post here on your behalf -- it might contribute more to the
| conversation.
| escapedmoose wrote:
| To each his own. I find the meandering prose refreshing.
| Especially in an article like this where I'm predisposed to
| find the subject distasteful, it breaks down the defenses a bit
| and helps me to keep an open mind. When you get to that point,
| you find it's not just about "how some TikTok manager games the
| system."
|
| Not that every article written in this way is a winner. But
| this one benefitted from the format imo
| TeMPOraL wrote:
| Funny how it had the opposite effect on me. Sure, I already
| think the whole thing is distasteful (to put it mildly), but
| all that "meandering prose" only drove it home to me,
| paragraph by paragraph, just how despicable the whole thing
| is.
|
| "How some TikTok manager games the system" is an abstract
| thing. A deep dive into a real step-by-step approach of
| pumping phony "content" and sneaking it past viewers'
| defenses - that's a concrete thing that couldn't be more
| disgraceful, save only for if the article explained that it's
| all done for a singular reason: every part of this self-
| referential loop of "content" is an opportunity to get
| viewers to spend money. That's where the funds for all of
| this come from: everyone there is a piece of the advertising
| machine.
| benreesman wrote:
| If you'll pardon my francais: this guy fucks.
|
| For all its obvious problems and downsides, this is how things
| are marketed and sold now: washing machines, (videos of)
| blowjobs, record deals, political candidates.
|
| Two examples I would cite are Tim Henson of Polyphia (who might
| be the first complete natural in this world, he's like 26), and
| Dove Cameron, who went from epsilon to 8 figure followers in as
| many months as it took Kim years.
|
| Henson makes more from sponsorships from social, and more from
| social than from Spotify. You can't buy the two hottest guitars
| in the world right now without paying him like 9 different ways.
|
| Cameron is making a run at TSwift: the ALCHEMICAL project is
| already a feeding frenzy without a track dropped, she's had a lot
| of vocal coaching since "Taste of You", she's got REZZ and Diplo
| and apparently Sean Penn in her corner, and has more style in her
| manicure than Swift's team was able to muster for the entire Eras
| tour. And frankly, I'm rooting for her: I miss white people pop
| music with some style.
|
| Someone is going to make a metric fuckton of cash implementing
| bandits and low-temp LLMs and hashtag embeddings working _for the
| creator_ rather than as part of the "the Algorithm".
|
| I've failed to interest any investors in this even as one of the
| main people who "worked on The Algorithm", thanks for nothing
| JPow, but someone will make a mint.
| uxcolumbo wrote:
| Why is this being downvoted? Because of the swear words?
| rchaud wrote:
| The future of music is here, and as this article lays out, it is
| incredibly bleak.
|
| > Flooding the zone this way caused TikTok's algorithm to funnel
| posts featuring "Toxic" into the feeds of users who consume gym-
| centric content [... ]they began to weave YoungX777's clip into
| videos targeting related subcultures--like the region of TikTok
| obsessed with highlights of soccer players bursting past hapless
| defenders.
|
| It's a well-known joke that soccer highlight compilations have
| some of the worst-sounding music known to man, an alternate
| universe where vocal trance and dubstep never fell out fashion.
| PartiallyTyped wrote:
| > vocal trance
|
| I find this comment offensive.
| martin_a wrote:
| I miss mid 2000s vocal trance. Let's see what Spotify has to
| offer on that.
| gopher_space wrote:
| Sade, dis-moi?
| xyst wrote:
| Reminds me of how clickbait farms were raking in the
| advertisement cash from Facebook during the height of its usage.
| Then FB updated the algo and all of these clickbait farms went to
| zero.
| anon84873628 wrote:
| Interesting article to learn what's up with "kids these days".
| And so this was also my favorite part:
|
| >When the Ovrthro call was done, I half-jokingly noted to Magana
| that the collaboration between Langevin and SyKo sounded like a
| prime example of synergy. He said he'd never heard that word
| before but that he loved it and would be incorporating it into
| his recruitment spiels from now on.
___________________________________________________________________
(page generated 2023-10-28 23:01 UTC)