[HN Gopher] Who will amplify this and why? (2021)
___________________________________________________________________
Who will amplify this and why? (2021)
Author : jger15
Score : 37 points
Date : 2023-04-04 12:11 UTC (1 days ago)
(HTM) web link (sparktoro.com)
(TXT) w3m dump (sparktoro.com)
| photochemsyn wrote:
| These are some circumlocutious explanations on how and why
| platform owners have tended to implemen recommendation algorithms
| that drive content to their preferred creators:
|
| > "They're driven by incentives to earn engagement and
| amplification on their own platforms. This means they're unlikely
| to amplify, reference, or publish your work unless there's an
| obvious connection to that goal."
|
| Translation: Their revenue is driven largely by advertising
| revenue, and upsetting advertisers is bad for the bottom line.
| Issue: how consolidated is the advertising market? They're
| selling services. So how many buyers are there? If there's just
| one buyer, that's called monopsony, and the buyer can exercise
| all kinds of control.
|
| > "They have audiences with affinities and preferences of their
| own, and unless your work resonates with _both_ the amplifier and
| their audience, you'll struggle to get the share. "
|
| As we've seen, a major interested audience is the local
| authoritarian state power, which wants to amplify state-produced
| propaganda while censoring a wide variety of content it views as
| inimical to its interests. This is not just a theme in China, as
| the current RESTRICT Act push by Congress and the White House
| demonstrates.
|
| Now, if the corporate interests buying advertising are joined at
| the hip to the government authority pushing propaganda and
| censorship, well... the only content that will get amplified is
| the bland neutral drivel that doesn't displease either party.
|
| Sure, there's still a market for high-quality content that's
| driven by popular audience interest - but it's getting to be a
| bit narrow in scope, isn't it? More and more topics have become
| tripwires for cancellation bombs.
| urbandw311er wrote:
| Excellent piece for content creators wanting more eyeballs but
| perhaps a little naive on what triggers virality (hint: sadly
| it's not a meritocracy)
| ignite wrote:
| I believe clickbait titles are frowned on in Hacker News. Please
| edit title to something that speaks to your subject matter.
| sophacles wrote:
| Its literally the point of the article, as well as it's title.
| Perhaps you should try reading the material - It's clearly a
| title about marketing, and looking at the article, it's about
| marketing.
| fsckboy wrote:
| I'm not taking either of your sides in this mini-debate, but
| I clicked on this article because I thought it would be about
| amplifiers, electronic circuits, for a specialized purpose
| musicale wrote:
| > Perhaps you should try reading the material
|
| I like this HN guideline:
|
| > "Did you even read the article? It mentions that" can be
| shortened to "The article mentions that".
| pjc50 wrote:
| It's a great explanation of a fairly basic point: in the internet
| marketing era, virality matters more than how good the thing
| you're marketing actually is, which means you need to trip the
| emotional engagement hooks to get people to hit the retweet
| button (or other platform equivalent).
|
| This has all sorts of negative consequences for what content gets
| produced.
| gravitate wrote:
| > virality matters more than how good the thing you're
| marketing actually is
|
| And people have reverse engineered how viral content on the
| Internet spreads, and there's entire playbooks on how content
| can go viral. Things like:
|
| - Is it short and quotable?
|
| - Does it trigger an emotional reaction?
|
| - Is it recent or related to current events?
|
| - Does it spark outrage?
|
| - Is it novel?
|
| - Is the author credible and do they have a large following?
|
| Basically any viral tweet are all those.
| urbandw311er wrote:
| Or also many of the points made extensively in the actual
| linked article.
| musicale wrote:
| > - Does it spark outrage?
|
| Feeding the outrage machine seems to work well.
| JALTU wrote:
| AI cat videos for the win.
| tomcam wrote:
| I knew you mean chicken videos, but whatevs.
| hammock wrote:
| And by virality you mean well-tuned for the algorithm.
|
| This issue seems like one that Google (or search more broadly)
| addressed early on, and spared no expense at making sure their
| pagerank algo surfaced quality pages no matter how people tried
| to game it.
|
| It doesn't seem that social media platforms have taken the same
| approach with their algorithms, for one reason or another
| pjc50 wrote:
| Quality of the _content_ is not the concern of platforms;
| they only want to maximize eyeballs for the ads. Whereas
| google, incredibly, had to establish itself as better than
| the competition at providing quality pages.
| hammock wrote:
| Yeah, what is the insight though as to why the difference?
| A social platform and Google both rely on active users,
| visits and time on site. Both make their money by selling
| ads. Both risk losing users if their algorithm sucks. Yet
| they tune their algorithms differently.
|
| On reflection, two factors come to mind:
|
| 1) Social content is inherently emotional - people come to
| the platform get a dose or three of dopamine. Search is
| unemotional in that way, people are only looking for
| accurate information. In the way that Wikipedia straddles
| the line between dopamine-driven content and search
| information, I wonder what Wikipedia would look like if it
| had ads on it?
|
| 2) Social ads are built for reach and impressions, not
| clicks (this is changing though), while search ads are
| built for clicks off-platform
___________________________________________________________________
(page generated 2023-04-05 23:02 UTC)