The dead internet vs the Lightning network \ stacker news ~mostly

pull down to refresh

The dead internet vs the Lightning network

3585 sats \ 33 comments \ @elvismercury 16 Jan mostly_harmless

The dead internet theory has been around for a while. In broad strokes, the idea is that for years -- since 2016-ish -- the content on the accessible internet has been mostly bullshit, generated by sock puppets or bag-holders or foreign powers or bots, and of course these groups are not mutually exclusive.

This has been on my mind for a while, and I'll have more to say about it later. But I read something today in Zvi's excellent Substack about the intersection of DIT and gen AI that shook something loose in my mind:

It is taking longer than I expected for this type of tool to emerge, but it is coming. This is a classic situation where various frictions were preserving our ability to have nice things like Reddit. Without those frictions, we are going to need new ones. Verified identity or paid skin in the game, in some form, is the likely outcome.

Basically, DIT has probably been nonsense, but it will probably stop being nonsense pretty quick because the tools for ruining the internet got a massive upgrade and people will eventually figure that out at scale. We've already got a taste of this on SN, in fact, with bot behavior being given as a prominant reason for muting stackers in a recent discussion.

Maybe SN is on both tails of this one -- a more likely target for bot activity (owing to the chance to make real money here) but also possessing more tools to fight it with (owing to the costly signals required to participate). It provokes reflection on what is valuable about online interactions in the first place.

If bullshit bot engagement isn't what we want, then what do we want, and how can we get it?

view all related items

673 sats \ 2 replies \ @freetx 16 Jan

The content on the internet is made up of 2 main factions: DIT and Insane People.

99.9% of Wikipedia users only browse and never create
97%+ of redditors only browse and never create
99.9% of youtubers browse and never create
same general level is probably true for X, Facebook, Instagram, etc....

Critically, there are power laws which govern those who create. Lets take wikipedia as an example: Wikipedia has about 110K active editors and gets about 114M daily active readers (so creators only make up .1% of total audience). However out of the active 110K editors, 1000 of them create 75% of all content!

Further power-laws exist within that group of 1000. One of the top editors, does 385 edits per day and has been doing so for 15+ years. Thats one edit every 4 minutes (assuming he never sleeps)!

The top reviewer on Amazon has 35K reviews, averaging over 3000 per year (8 per day).

The same power-laws are true for reddit, youtube, instagram, etc.

Conclusion: Most of what you read on the internet is either a bot or an insane person.

44 sats \ 0 replies \ @028559d218 16 Jan

100% agree with the 'dead internet' theory. Twitter is imo 70-80% bots... and Youtube is getting worse.

We should be prepared for a future where... without the incentives of Bitcoin and Lightning specifically 70% of the internet is... bot comments and 'engagement'.

I would love to see just for one day where all the youtube 'comments' and engagement is actual people. I think it would look vastly different.

11 sats \ 0 replies \ @Bell_curve 16 Jan

a bot or an insane person

I was wondering who writes 3000 reviews per year

I tried writing an Amazon review once. I didn't meet the minimum character requirement or something. That was the end of my noble experiment

800 sats \ 7 replies \ @justin_shocknet 16 Jan

bullshit bot engagement

Problem is that's exactly what people want, to refresh and have new stuff appear. DIT isn't just a theory, it's what we've created through incentives.

Reality is that without bots and bullshit, everything is pretty stale relative to how many hours people are on their phones looking at shit.

Most of the Elon derangement syndrome that gets posted here is pretty funny because if you actually pay for the service to filter out the bots and curate your follows, it has become exactly like Nostr... a ghost town of stale feeds and low-signal dopamine seeking boredom posts.

SN is distinct from things with a feed because its more akin to the classic discussion board... I'm willing to bet the stats for desktop systems (keyboard) are over represented vs mobile devices relative to feed based apps.

Mobile devices are low-quality internet time, that results in low quality social apps. More productive devices like desktops produce higher quality content, but at lower quantity correlated to mobile/desktop device time.

2016-ish

Tipping point of mobile vs. desktop

84 sats \ 0 replies \ @MaxAWebster 16 Jan

Mobile devices are low-quality internet time, that results in low quality social apps. More productive devices like desktops produce higher quality content, but at lower quantity correlated to mobile/desktop device time.

This is so true. I recently removed most apps from my phone (including messaging). I still have a browser but may cut that next. So far, so good. My laptop time is much more focused and much higher quality than mobile time. And SN gets the majority of it...

100 sats \ 5 replies \ @elvismercury OP 16 Jan

SN is distinct from things with a feed because its more akin to the classic discussion board.

This is an interesting observation. The underlying paradigm is important, and often the hook gets swallowed without anybody thinking much about it, e.g., we all "know" that online social sites have a feed, and so a feed seems like the obvious thing to do. But every choice exerts force in the world, and the way of being in response to a feed-like-thing is different than the way of being in response to something else.

It's a worthy exercise to think of what the desired outcome is, and then try to realize that outcome through conscious design choices, incentives, etc. Easier said than done, of course. I'm in a proximate industry and am stunned by the gravity of default choices on thinkable thoughts.

218 sats \ 4 replies \ @justin_shocknet 16 Jan

I like Elon's stated objective of "unregretted user minutes", that's a distinction from user minutes more generally which has been the standard, and has long used tricks to keep people addicted to serve more ads.

I think unregretted user minutes is an evolution to the times, because people are going to spend the minutes anyway... the arb is now where those minutes aren't regretted and how they're monetized.

So, we have the feed as the emergent standard since when, MySpace era probably? And feeds have always been twitters whole thing, true also of facebook and so on. Feeds being a lindy pattern, the only option for them is to improve incrementally because we don't know what the successor to feeds looks like yet.

To make matters worse, feeds only have 2 properties to compete for minutes on, network and algorithm. Yet, both of these inevitably reward bot and bullshit DIT behavior because feeds don't work if they're stale.

The SN/Reddit-like social filtering system is the only alternative I can think of where newness matters somewhat less than in feeds, but is probably still more ephemeral than whatever ultimately starts taking user minutes away from feed based apps.

208 sats \ 3 replies \ @elvismercury OP 16 Jan

I like Elon's stated objective of "unregretted user minutes

I hadn't heard that phrasing, but it's a noble aspiration for sure. "Regret minimization" is a storied concept from decision theory, seems worth applying to this new universe.

The SN/Reddit-like social filtering system is the only alternative I can think of where newness matters somewhat less than in feeds, but is probably still more ephemeral than whatever ultimately starts taking user minutes away from feed based apps.

I am titillated by the idea of breaking the "newness" thing out entirely -- this is the source of my commitment to evergreen-ness but I admit that there's not much of a model for it at anything like scale, at least as far as I've seen.

The distance between some non-newness-based interaction paradigm and the current state of things seems both daunting, and a substantial opportunity. Nobody whose current existence is underwritten by a traditional business model will go after it, that's for sure.

54 sats \ 2 replies \ @justin_shocknet 16 Jan

evergreen-ness

Framing of that may need to change, less about what's absolutely new and more what's newly relevant to the user

How often do you get a recommendation for an old movie or TV show? Read a wikipedia article that's not been updated in months or years? Peruse an old review or blog post?

There's an infinite amount of content that would be stale on a feed, but you haven't seen yet because you didn't seek it out probably because it hasn't been relevant to you... so whatever replaces feeds (or a new algo for feeds) assesses what's new with you and your mind and not so much what's been newly posted.

0 sats \ 1 reply \ @elvismercury OP 16 Jan

You described evergreen the way I intend the term -- doesn't matter when it was written, it's relevant now. Although you added a distinction that complicates it a bit -- the relevance would evolve based on your own evolution.

Crime and Punishment references from a hundred years ago were not relevant a month ago, now they are. So it's less about the thing, even if it's a high-quality and timeless thing, and more about the interface between you and the thing, whoever you happen to be at the moment.

Maybe the science fiction version would be an intent-less / ambient search engine.

27 sats \ 0 replies \ @justin_shocknet 16 Jan

Maybe the science fiction version would be an intent-less / ambient search engine.

I think that's what Google at one point said they wanted to do, it just requires a lot of your private data... ironically the same people that would probably most enjoy higher signal feeds are going to be more stingy with their data than average feed dopamine enjoooyer.

Surely a fortune to be made figuring this out.

212 sats \ 2 replies \ @k00b 16 Jan

Maybe SN is on both tails of this one -- a more likely target for bot activity (owing to the chance to make real money here) but also possessing more tools to fight it with (owing to the costly signals required to participate).

I've been thinking about this a lot too. My hope is the money means we get the better bots.

199 sats \ 1 reply \ @CheezeGrater 17 Jan

"better bots" can also mean better at faking being human, and better at farming engagement though.

The cost to post is effectively an evolutionary pressure.

200 sats \ 0 replies \ @elvismercury OP 17 Jan

I'm guessing @k00b is making a deeper philosophical statement -- at some point, if bots got really good, you wouldn't mind that they were bots. They would be valuable members of the SN community.

It's not that hard to imagine that, actually, but they would be different bots than the ones we've been seeing to date.

205 sats \ 5 replies \ @charonnakamoto 16 Jan

I am a new user of SN(joined in Jan 2025 actively), and paying to post is most certainly a feature. The signal-to-noise ratio here is incredible. I hardly spend time on X anymore, which is full of engagement farming. I think all platforms can gain from this model of pay to post/reply and improve their Daily active users but will be disincentivized as that will eat into their advertiser led model.

114 sats \ 4 replies \ @028559d218 16 Jan

Proton mail and their 'pay-to-use' model, which has been extremely successful, is a great example of making a 'free' service (that's not actually free) much higher quality through pay-to-post.

SN is the quality place where it is due to 'pay-to-post'... people think before posting if it costs them 10 cents. Youtube should be the same.

So should social media in my opinion. The costs would be minimal... but the quality of content would be so much higher.

54 sats \ 0 replies \ @MaxAWebster 16 Jan

Yes. Same with Kagi for search.

There's a large and growing contingent of us who want transparent cost/value.

0 sats \ 2 replies \ @elvismercury OP 16 Jan

Say more about how the Proton model maps to this. I'm not connecting it.

239 sats \ 1 reply \ @028559d218 16 Jan

Google is 'free'. Hotmail is 'free'. The 'cloud storage' the 'email service' the 'calendar' the apps associated... all these features are 'free' at Google and Microsoft.

But they're not really. Those services have to be paid for somehow and maintained... and so google users are NOT the customers. Google users are the product and their data + privacy is constantly being sold to advertisers to generate revenue. Google is not an 'internet' company it is a massive advertising company and "you" the user are the product.

Facebook is another horrible, hideous example of this. Noone 'pays' to use facebook... and yet they make money big time. How? By selling user data and privacy to advertisers and AI companies. Facebook totally, completely owns all your data and all interaction on the site to monetize it.

Twitter is another example of this it's 100% the same. It makes money driving 'engagement' to advertisers that's it full stop. Anything to result in 'engagement' (even total nonsense) is what they promote.

Years ago, Proton Mail realized that asking people to 'pay' for something 'free' (email) could protect their privacy and result in a better service... because user data doesn't have to be sold to advertisers.

It completely, totally fixes the incentives of online services and engagement. All the broken incentives, lack of privacy, and underhanded advertising is removed. You get what you pay for and beyond that Proton has taken Bitcoin payments for years... so that they don't get cancelled by the credit card companies for selling encrypted/private communication government doesn't like.

Personally I believe that Lightning is long-term, enormously necessary to be 'anti-bot'. If bots on Youtube or Twitter bad to pay 1 cent (on lightning of course) per comment... 90% of them would disappear overnight. Bots and spammers rely on volume and 1 cent a comment wouldn't effect legitimate users... but spammers won't pay 100s or 1000s of dollars to spam comments with very low ROI. I think it's the future and increasingly necessary to 'clean' the internet of total nonsense.

Imagine if creating a twitter profile cost 10 cents.. and maybe 1 cent a comment to post. What % of profiles would disappear and how many comments would fade away? My guess is that 80% of 'interaction' would disappear as it's artificial engagement or bots. But it would be a better platform in the long run.

0 sats \ 0 replies \ @elvismercury OP 16 Jan

Good points, and well-said.

102 sats \ 4 replies \ @NovaRift 16 Jan

I never imagined there would be a platform out there on the Internet that has a completely different structure (pay for posting). I was a heavy X user 2 weeks ago; now I see how the internet really feels when the algorithm doesn't favor those who have a Bluetick only. No bots, nothing, only quality content. Thanks @k00b and @ek for creating Stacker News. Facebook is already a living example of the Dead Internet Theory, and soon X is becoming one. It's all engagement farming out there.

21 sats \ 2 replies \ @028559d218 16 Jan

Bots talking to bots.

0 sats \ 1 reply \ @oklar 17 Jan

Bots zapping bots?

7 sats \ 0 replies \ @028559d218 17 Jan

Bots don't zap.

0 sats \ 0 replies \ @nym 17 Jan

Welcome! A lot more real people here

54 sats \ 1 reply \ @Bit_Alb 16 Jan

If bullshit bot engagement isn't what we want, then what do we want, and how can we get it?

I think many stackers (if not all) will agree that “bot engagement isn't what we want”, but do you really think that all stackers (or the majority of them) can agree with each other on what is that “we want” and “how to get it”?

0 sats \ 0 replies \ @elvismercury OP 16 Jan

Almost certainly not, but it would be interesting to hear the desires made explicit and surfaced. Usually stuff like this proceeds without any intentionality, and then you just get a ton of complaining about it later.

269 sats \ 0 replies \ @Scoresby 16 Jan

@Darthcoin is right: SN's great innovation is you get to pay to post.

31 sats \ 0 replies \ @e0ce06e368 16 Jan

The "dead internet" is just fiat internet bloated with fake engagement, bots, and useless noise. The Lightning Network? Pure signal. Real people, real value, real skin in the game. Bots can’t fake sats. Pay to play filters out the noise. This is how the internet should be. Fix the money, fix the web.

0 sats \ 0 replies \ @ken 17 Jan

Incorrect API Key. Please try again

0 sats \ 0 replies \ @Taft 16 Jan

The AI machines can now make words, good words that look real. In the past, if you wanted to fool people online, you had to do the work. Now the machines do the job.

0 sats \ 0 replies \ @LowK3y19 16 Jan

Internet has been dead lol

0 sats \ 0 replies \ @nym 16 Jan

Yes, it will get exponentially worse in the coming years. I'm not sure of any solutions.