[lastofthemojito]

I feel like I've thought a lot about ads on the internet recently, and I hate the way things are, but I'm no closer to a realistic overall solution.

I've used an ad blocker for a while because I'm tech savvy and I know I can easily have a better experience that way. I was too late installing an ad blocker on my father's computer - he was scammed out of some money when he mistook a scammy ad for a legitimate message to call his bank.

Due to the experiences of countless others like my father, the FBI now recommends folks run an ad blocker.

But if everyone did, then the whole thing falls apart right? If Google or Apple built a great ad blocker into their browser and shipped it with the next update, content creators would respond pretty damn quick right? Everyone from SMBC to the NY Times would put all of their content behind a paywall, or in an app, or something else that's worse than the status quo right?

sigh

[ThrowawayR2]

The attention economy is powered by payments from maximizing ad impressions but the public is increasingly recognizing that the attention economy is mentally and culturally toxic. To break free of the attention economy death spiral that we're in, internet advertising has to undergo a pretty drastic dieback. In other words, the whole thing has to fall apart and the sooner the better.

[lazide]

That would mean the death of at a minimum Meta (market cap $1.5 trln), most of Google (market cap $2.1 trln, lets say half would die). So what, a couple trillion haircut on the US stock market, and more of a GDP hit than if the entire US military suddenly stopped getting funded.

Not to mention the secondary effects - a LOT of businesses are built based on online ads. A LOT.

[ben_w]

> That would mean the death of at a minimum Meta (market cap $1.5 trln), most of Google (market cap $2.1 trln, lets say half would die).

And nothing of value was lost.

https://www.youtube.com/watch?v=jrFv1O4dbqY

https://www.youtube.com/watch?v=uSGVk2KVokQ

> So what, a couple trillion haircut on the US stock market, and more of a GDP hit than if the entire US military suddenly stopped getting funded.

Not even.

Market Cap is more like net worth than like GDP, GDP is more like annual revenue (both are still somewhat different, hence "more like") — Meta's global revenue is $164.5 billion, Alphabet's is $350.02 billion, so even if that's entirely in the USA (it isn't), and the money spent was purely positive-sum (it isn't), even combined that's less than the US military budget of $849.8 billion.

[lazide]

I guarantee you if Meta all the sudden couldn’t sell any ads tomorrow, it would be a lot more than a 1 trln GDP hit we’d experience.

[wizzwizz4]

If we'd lose so much GDP, without losing anything of value, perhaps GDP is not a useful measure of the economy.

[lazide]

Easy to say when you don’t have bills to pay.

[relaxing]

I think there are other advertising channels that could pick up the slack.

[tsimionescu]

Both Google and Meta could just charge for using their services. I already pay meta to see no ads. So there is no reason that they would entirely die.

[ThrowawayR2]

CDOs, the financial instrument behind the 2008 financial collapse, had a tremendous amount of market value too. A "LOT"(sic) of businesses were built on selling tobacco products, lead based gasoline, and high sugar ultra-processed foods too. Perhaps you should rethink your argument?

Creating economic activity doesn't justify harming people and it has become thunderingly obvious that titillating and outraging viewers constantly to maximize ad impressions is harming the public.

[lazide]

I’m saying it’s easy to say until the consequences arrive. And with something this big, the consequences will be big.

But then I’ve lived through 3-4 major downturns so far - and seen the ‘oh shit’s land.

[ThrowawayR2]

I've lived through 3-4 major downturns too and, yeah, it's hard. Like discontinuing leaded gasoline, we either pay the price now or pay much worse later.

[lazide]

I don’t disagree, I just cringe when I see (presumably young/haven’t been through it before) folks cheer for destruction.

We had alternatives to those things, when we banned them, after all. Or didn’t ban them at all, just taxed them more.

[pixl97]

The best day to stop smoking was yesterday, but stopping today is the second best.

The consequences of addiction are huge, and the longer you stay addicted the worse it gets.

[lproven]

"You say that like it's a bad thing."

[UltraSane]

no pain no gain

[NoMoreNicksLeft]

Google and Meta collectively employ how many software people in the United States? The jobs hit would be unimaginable. High octane nightmare fuel, that idea.

[michaelt]

Doesn’t seem like a nightmare to me - a trillion dollars spent on ads that get blocked is a trillion dollars wasted.

The real nightmare is hundreds of thousands of people on six figure salaries wasting their lives producing ads everyone hates.

[NoMoreNicksLeft]

You're right. What was I thinking? Instead, they should be unemployed and starving, and all the services they bought from other Americans with their $200k/year salaries, those Americans should also be unemployed and starving too. All because you don't want to see an ad on whatever ridiculous social media it is that people are visiting now days. That will make the world right.

[relaxing]

Get a grip, mate.

[ThrowawayR2]

"...total advertising has averaged about 1.3% to 2% of U.S. gross domestic product (GDP) over the past 20 years. ... In 2015, digital advertising corresponded to about 50% of total advertising spending and more recent estimates point to a share closer to 65% of total advertising spending. That represents about 0.6% to 1.1% of GDP. ..." from https://www.stlouisfed.org/on-the-economy/2024/oct/rise-digi...

So, nah. It's significant but not "nightmare fuel" even if online advertising entirely went away. (Which nobody is proposing, by the way.)

[relaxing]

Google AI summary says:

> As of December 2024, Meta Platforms had 74,067 full-time employees. This is down from 67,317 in 2023

lol. At least Gemini doesn’t make the same mistake for its makers:

> As of 2024, Alphabet, Google's parent company, has 183,323 full-time employees

Of course that’s total headcount, I couldn’t find figures for just US based SWEs.

And in the US alone there are something like 200-400k SWE+IT job openings and 7M openings across all sectors.

Verdict: after the dust settles, it would be a net benefit for everyone to get those bright minds working on real problems.

[derkster]

I'm old enough to remember an internet before non-porn ads. Being Hacker News, does someone have any hard idea on infrastructure costs in the late 1990s compared to now?

In 1997, when America Online went from an hourly rate to unlimited hours for $20/month. Besides adult ads, I remember a lot of "work from home/get rich quick" ads - and that's all I can personally remember. Almost every forum and community I can remember were serious labors of love, especially the early "hacker" communities. But, we lost that internet a long time ago. I, personally, would love for an internet that could emulate those early years of being able to log in, communicate, and make friends with random, sometimes sketchy, weirdos, all within 45 seconds of dialing up. There weren't algorithms that decided who was important and who wasn't, you just had your words and wits.

That was a tangent, but my point is - would the internet improve or decline if there was a serious "adpocalypse"?

[stavros]

I wonder how cheap a forum is to run now. It used to be that you could run a janky thing on nothing hardware and people would still use it and find value, but now even just running the software is much more demanding and expensive. Never mind the fact that the UI has to be slick or people will be put off, because the bar is much higher now.

These factors have made it less economical to just run a popular forum out of your own pocket, I think, so the barrier to entry is higher.

[ChrisMarshallNY]

Cheap isn’t the relevant metric. I’ve run a couple of forums, over the years. It’s entirely possible to customize the software, to make the user experience, for both users and administrators, quite good.

Moderating and managing the forum is the pain.

Also, it can be fairly corrosive to personal mental health.

I recently had an old acquaintance reach out to me, asking me to revive one of my old communities, that was killed by Facebook. This person argued that many of the old community members were now thoroughly sick of Facebook, and would want to go back to the dedicated forum.

I politely declined, and was surprised at how the idea really horrified me.

[rat87]

> but now even just running the software is much more demanding and expensive. Never mind the fact that the UI has to be slick or people will be put off, because the bar is much higher now.

I'm skeptical of this. There's likely some open source software that's good enough and considering the advances in hardware likely much cheaper to run then it was way back when.

Besides the fact that many people prefer large platforms like reddit or Facebook or discord where they already have an account (don't have to make another one) I'd say the problem is the human one not software. Time and effort to popularize the forum and either paying for moderation or convincing people to volunteer to moderate and deal with inevitable shitshow that comes with moderating controversial topics even if forum isn't really related to them.

[TylerE]

Good enough isn’t nearly good enough. You’re competing against apps, offering just a much better use experience.

One bit of feedback I got repeatedly from users on a couple communities (now dead) I helped run was how much harder posting image and video content in a forum is - true user generated content I mean, not shared from YouTube. Even that meant ricking around with bbcode.

[homarp]

i somehow envision a forum as text first. Image and video hosting is 'out of scope'.

[TylerE]

Times change. About 95% of the world’s population now carries a good-to-amazing camera in their pocket almost all the time.

[jrmg]

Yeah, it should not be hard! Cheap hardware is orders of magnitude faster than it was in the 2000s, and we could run dynamic forums written in janky inefficient Perl on it, and serve thousand's of requests a second.

[superkuh]

I've been playing NationStates(.net) for 20+ years now. They run a php forum as part of the civic simulation php site. It hasn't really changed in the 20 years. Lately the entire site has become completely inaccessible to many of us because they've deployed a cloudflare block. They say they had to do this because the increase in bot traffic was slowing down the site so much it was costing them money: https://forum.nationstates.net/viewtopic.php?f=15&t=562415

I never once noticed the site was slow. But now that cloudflare is blocking me and my non-corporate browser completely, no matter how many captchas I complete, it is very, very slow (dead).

It's shocking to me as someone that runs multiple websites (for the last 20+ years) that they're having so much trouble suddenly. I too see bot traffic on my sites increasing (ie, it's like 4:1 now instead of 1:1 bots:humans) but it is not a problem.

I have to wonder how much of this is a socially contagious hysteria. Computational resources and bandwidth have become massively less expensive over those 20 years. I get the feeling people are "having" to block bots from accessing their websites just because they "feel" they have to block them. They actually don't.

[Macha]

> Computational resources and bandwidth have become massively less expensive over those 20 years.

Your budget tier VPS provider on the other hand has only really doubled what you get over the last 10 years, and user feature expectations and AI scrapers really have caused resource usage to more than double.

The cost efficiency has really only benefited the middle end here really. HN is probably cheaper to run than it was 10 years ago. But the large sites are on cloud providers who have provided added services that you might not have used 10 years ago to keep up their margins, and the intro tier VPS that "Bob's Friendly phpBB Forum" might be on isn't getting you much more, which matters when Bob's revenue is $60/year from a handful of the most investigated regulars only.

[kibwen]

Holy hell, I can't believe NationStates is still kicking, or that Firefox has managed to hold onto my login details for 20 years. Truly a more innocent age that we'll never be able to return to.

[multjoy]

I remember running a semi-popular phpBB back in the early 00's. There was nothing like Digital Ocean or Linode. You can run a small instance for pocket change today.

[Kye]

There were VPSes and managed servers back then, but most people went with shared hosts that supported PHP and MySQL. That's how WordPress got so big.

[a_shovel]

I've heard legends of a time, perhaps back in the mid 2000s, when companies paid so much for banner ads that you didn't need another revenue source for your website with Flash games or what have you. If such a time ever existed, it ended long before Flash did.

[TylerE]

It ended when the vc money in the original dotcom boom dried up.

[cameronh90]

I might be unusual, but I’d be quite happy to pay more than my ad value (less than a cent per pageview in most cases) to sites, but there’s no system to do so.

There’s zero chance I’m going to subscribe to every site that I might hit a few times a year.

[trilbyglens]

Everyone just assumes that the only way to do ads online are from these giant, shitty ad platforms like Facebook or Google. When really, you can have smaller ad networks that provide curated and content tailored advertising. It's just a lot more labor intensive and it doesn't scale to nation-state levels of money like Google Ads does.

If sites just embedded their own ads, then no ad blocker would be able to catch that at all because there would be no way to identify them separately from the content.

[stavros]

A smaller company like Hiveworks, for example?

[ccppurcell]

I think you're wrong about that last part. For ads to make sense as a revenue stream you have to basically auction them, for example at different times and different countries etc. So there has to be some automation, you can't just manually include each ad. This will have to be done programmatically, and there will always be traces (e.g. in div classes). You can try to randomise these or what have you, but it's basically a cat and mouse game with ad blockers.

[TRiG_Ireland]

Their used to be a system (which was created by a webcomic artist, though I forget which one) where ads could be sold in slots of time, completely independent of clicks. But it died a few years ago.

[piperswe]

RIP Project Wonderful, I loved that it was an ad network that respected privacy

[elsonrodriguez]

Dinosaur Comics I believe.

[JBiserkov]

> there would be no way to identify them separately from the content

We'll use AI

[dmurray]

> Everyone from SMBC to the NY Times would put all of their content behind a paywall, or in an app, or something else that's worse than the status quo right?

The NYT already puts its content behind a pay wall, and SMBC (as of this update) doesn't rely on third party ads either. So it's not clear that something else would be worse than the status quo for them.

Some content creators would certainly suffer, but if online ads were not normalized, others would thrive.

[seanalltogether]

The answer to no more ads is you can't be independent. You have to have service aggregation.

The fundamental problem is that you can't grow an audience all your content is behind a paywall. So you have to offer some content for free, but if you're an independent comic artist like smbc, no one is going to visit your site if only 1 comic per week is free.

But if a lot of creators are behind a single paywall, and that paywall was able to feature 10 random free webcomics every day, then you pull in lots of visitors and have more opportunity to convert them into paying customers.

Now the harder question, how do you convince a bunch of independent content creators to band together behind a single paywall.

[hakfoo]

I think in a way, we already sort of have precedent for this, in the comics sector.

The Japanese-style "phone book sized" manga weeklies/monthlies are the manifestation of this idea. Technically it's not a "free" teaser, but if you buy the issue for one or two "known" series you want to try, the rest is shoveled into your hand for free.

It also provides a good stickiness factor-- if you originally bought the magazine for one or two series, hopefully you've glommed onto others by the time those end or you lose interest in them.

[throwanem]

Nebula did it.

[TRiG_Ireland]

Nebula is mostly video, but also hosts a bunch of podcasts. I wonder whether they could be persuaded to expand their range of offerings to other media.

[amanaplanacanal]

I'm not entirely sure it would be worse than the status quo. If all the free sites were the ones run for the love of it, and all the paid sites were commercial garbage, that might actually be better.

[lou1306]

a) You can set up your own advertisement/sponsorship system and, unless some bloke took the time to specifically write rules targeting your website, the ads would not be blocked. Bonus points: this prevents most malicious ads, which are another major reason people use ad blockers, from appearing on your website, simply by virtue of you not letting them in.

b) You can always follow a combination of free (unpaywalled) content + premium content behind a paywall/Patreon/etc + merchandise + Kickstarter for more ambitious projects.