When you read something, why do you do it? Maybe you were looking for it. Maybe someone linked it to you. Or perhaps, it just happened to be there in front of you, and so you read it.
Of all the reasons to read something, that last one is unique. It doesn’t matter what it is, only where it is. This is why ads work.
Nobody is looking for an ad, and people generally don’t link and discuss ads between themselves (unless it’s for side-channel reasons). But if it happens to pop up in front of you, it doesn’t matter how little you care: the effort and time needed to figure out it’s an ad to begin with means you’ve already gotten the one glance most ads optimise for to begin with. The only way to escape this, at least in theory, is to block the ads. Or is it?
There is No Escape #
Ok so we’ve determined that since the ads are generated externally and inserted into pages based on a bunch of metrics with no further integration, we can just filter those out. Now we can enjoy the underlying content unblemished in its original form, without being distracted from what we’re actually trying to achieve. The writer is saying something, and I want to understand what they are communicating to me. The ads are just there because, well, they have to make a living, and they just wouldn’t be writing this thing were it not for that.
Anyway, turns out this algorithm to filter – or perhaps block – ads was too effective. Everyone’s unhappy! Sure traditional ads still make plenty of money, but a surprising amount of people do block them. And hey, we want a higher conversion rate anyway. Great news, I had an idea – what if we pay the person writing the content to integrate it inside? Now, if you want to gain access to what they have to say, you will consume the ad.
There is No Escape #
I’ve been observing how the quality of the content I’ve been interacting with has been reducing. Some are just barebones ads for what they’re talking about, while others are essentially nonsensical and filled with a bunch of links to other nonsense. How did I end up here again? Oh yeah it was at the top of the search results.
Ok, I see, how specific an article’s metadata is to the search determines how well it scores in the search results. But the actual content is irrelevant. So you just fill it up, make sure it contains as much extra metadata to make the crawler happy, and then… turns off ad blocker Oh yeah that explains a lot.
There is No Escape #
Ads are eating the world, and you physically cannot escape them. Anything that is done for a profit motive in online media is going to essentially run on ads. Anything that is not done for profit will similarly not try to push its way on top of all the charts (since you only really need that to maximise ad viewing rates). As such, everything you’re going to see (besides RSS and similar subscriptions to things you know are good) is going to be profit motive writing optimised for ads.
How is this optimisation done? Simple – pump out as much as you can, that gets as good a rating as it can, to then get the most ads views. If you’re lucky, instead, it’s someone that has something to say, kind of, but will then integrate the ad directly into the content.
With this lack of separation, it’s becoming impossible to avoid polluting your brain with garbage. There are still ways to avoid some of the integrated ads, like the sponsorblock extension… But that doesn’t help when the content itself is compromised in order to become a better empty vessel for the monetisation scheme. And if it’s not monetised, it dies, or you simply haven’t heard of it.
If you want to extract anything useful out of a subject you’re not already familiar with and know good sources for, but don’t want to reduce your cognitive capacity for someone’s benefit, you essentially have to read the article without reading it, and then use this new information to decide if it’s actually worth reading or not! Ah, yes, perfect, that’s a good use-case for AI, perhaps. Wait…
There is NO Escape #
Anyway so turns out AI is really really good at SEO. But really really bad at writing something that is a purposeful message (besides marketing gimmicks). So now you’re not reading something someone felt forced to say in order to put $5 on their plate at the end of the week. You’re reading something regurgitated by a thing that believes baby pigeons are small copies of pigeons and also are drones.
There is no good test for whether or not something is written in order to make money. You can at best guess at the lack of a monetisation scheme, but that still doesn’t help! Plenty of people and platforms write so that they can then write on their CV that they write so that they can then work for a content mill so they have professional experience so then…
The only good way to ensure that what you’re reading is good content is to read it. And then hope you remember that that one site was actually kind of maybe possibly useful. You’re not gonna remember, and then you have no way of searching through just those.
Not an Escape, but an Improvement #
You cannot escape the ads. But there can be a better way.
We already have indexing engines that work well with static sites, such as Pagefind (tinysearch also exists). Imagine if you could have a self-hostable search engine. Here’s what it would do:
So whenever you find a website where you go “oh, ok, that’s actually trustworthy, someone that truly knows what they’re talking about”, you would go and you’d add it to your search engine for a given topic. You could also specify a crawler depth, usually probably “1”, where the depth determines how many different domain names it’s allowed to recursively jump to while indexing. This then gets put into one big index for this entire topic.
Suddenly, your signal to noise ratio problem is gone. Basically everything that is there, is there because you found it. And you’ve already gone through the work of figuring out that, indeed, it is relevant. You could then share these findings just like sponsorblock shares sponsor segment timestamps.
But who’s gonna write it, eh?
It feels like a lot of work in search engines has been in “one big search engine for all the things”, to replace google. But google is the way it is precisely because they want to sell you things through their own damn platform. Topic-specific search indexes make far more sense, especially when you realise you do need to filter out the noise to begin with, before it ever makes it into the index.
Because while there may be no escape from the ad economy, that doesn’t mean we shouldn’t do anything about it.