Show HN: An open source extension to block large media brands from Google search (github.com)

75 points by petertsfn ↗ HN
Google sends 16 large media brands (588 individual brands) a combined 3 billion+ clicks per month. They rank first page on 85% of all searches.

They are inundating the web with things like subpar product recommendations, AI written listicles, and cookie cutter reviews.

I don't like that, so I made an open source extension that blocks all of them from search results. (I'm not a developer, and I made this with Claude Opus's help)

31 comments

[ 2.8 ms ] story [ 61.9 ms ] thread
admirable work, but this really should've been a userscript instead - crossbrowser by default, easier to maintain, easier to modify on the user's end (whitelisting individual sites by commenting them out in the userscript).
I think it should be a maintained list of ublacklist which is very useful extension to manage the search experience on search engines (including othet engines than google).
Ideally plugins for all the various search engine result pages
It blocks nhs.uk which is definitely not a large media brand.
It's funny because it's critical of sites for low quality "AI written listicles" but it's essentially a list of domains written by AI. And the result is there seems to be no consistent application of the principles expressed.

Why are we blocking amazon.co.uk and amazon.in? Why not block amazon.com by whatever that logic is? hulu.com gets a block, but why not netflix.com? Or why hulu.com in the first place? If there is a reason, it would be valuable to express that so people aren't just blindly blocking sites simply for being popular.

It's actually the _only_ page worth clicking on in most of the searches it appears in!
UBlacklist is great at blocking search results from various search engines. We could import domains from this list in that. As others mentioned, not all sites are correct here though.

You mentioned the article about 16 media companies. What would have been really useful if you made a table/list of each media company and it's domains.

When I read the Housefresh article, I googled all dotdash meredith sites and blocked them using UBlacklist. Didn't have enough energy to find the other 15 company sites.

The list seems like a great way to ruin your search experience, honestly. It's blocking things like apple.com, microsoft.com, samsung.com. If you're living a life completely free of their products, they're unlikely to appear in your search results, but if you are, you're replacing official content with blogspam and AI generated spam. Same goes with news, you're getting random sites instead of trustworthy ones.
In the previous [flagged] submission, the author provided some comments about this project: https://news.ycombinator.com/item?id=40569686

In any case, the title is "large media brands" so urls like "nytimes.com" and "washingtonpost.com" are consistent with that theme but the "blocked_domains.txt" also has "apple.com" and "microsoft.com" which are not media publications. If the motivation was avoiding SEO spam when searching for reviews of "air purifiers", domains from non-media-publishers like Apple.com and Microsoft.com are not part of that problem.

If instead the author is really trying to block "any big companies" rather than just "big media", it seems other giant non-media companies like NVIDIA, Toyota, JPMorgan, etc would also be on there.

(comment deleted)
(comment deleted)
Hello. Why it worked this way is because I can't find a list of 588 websites by thoae 16 companies (I've contacted the people in those articles to see if they have a list), I decided to find a list of top 1000 popular sites (which is much easier to find) and then try to clean those up. Obviously that's not a perfect method and leaves a lot of sites flagged by mistakes, but I didn't find any other way. I left it open source since I want people's help if they find mistakes in the list or want to supply their own list. There's nothing special about the code, I pretty much asked LLM to write it the whole way. I didn't set out to make money from this project, I just want more people to see what searches have become: a bunch of people with a lot of money pretty much own 90% of all searches.
> I decided to find a list of top 1000 popular sites (which is much easier to find) and then try to clean those up. Obviously that's not a perfect method and leaves a lot of sites flagged by mistakes, but I didn't find any other way.

This shows that your primary mistake is not that you’re not a developer, but that you are not willing to devote time and be analytical about the problem you want to solve.

Yeah, as someone who has built similar projects, this is a few hours of research to come up with the list of domains. You really should’ve done that; most of the comments here are negative, and ALL are related to the list you used.

It’s a cool idea and personally I would still do the research, revise the list, and then post it to HN once more. It’ll get a much better reception.

It's for Google Chrome. The idea is great, i'd like to see it for Firefox and search engines other than Google (DDG, …)
Google Chrome users probably dont care about privacy or Google snooping.
We are in a weird internet time where if you consume something online, it's there because somebody has spent time and money -- sometimes a lot of it -- to make it that way. This is true of both brands and individual content providers.

That doesn't make the content wrong or bad, but it means that the diverging incentives between producers and consumers is going to continue a drift towards easy, mediocre, engaging material and true deep-dive information, even with tech articles. The net went from "search, find the source, deep-dive, make tentative judgments" (like one would do at a library doing research) to "search, pop-in, see the listicle five-minute fact-bomb explainer vid, then assume you know everything important there is to know." (as one might do, say, browsing items as a yard sale).

If you're looking for true knowledge, you're going to have to deal with this more and more. It's a tend that's continuing, not some current state of affairs. If you're looking to be entertained and waste time playing games and absorbing random trivia and political justifications for things you already believe, you're in luck: a lot of people and a lot of money are hard at work to keep you happy.

I wonder how the decline in personal blogs and websites (who even starts one today, when you can just be popular on youtube/twitter/instagram), affects rankings due to lack of websites linking to other websites.
We've fallen into some kind of long-form desert.

I used to read programming books not for a laundry list of features, but to get inside the author's head and start thinking about programming the way they did. It's not as simple as a simple declarative statement. Many times the author themselves couldn't express the value I got. I don't see how you reduce that any and I'm not sure I'd want to consume it if you could.

Blogrolls and Webrings are largely extinct unfortunately.

Even for the big media sites, they mostly link to other pages on their own website.

> I'm not a developer, and I made this with Claude Opus's help

PSA: just clicked into this post after reading another submission complaining about how companies don't take security seriously enough.

If you care about the security of your browser, installing a dozen random chrome extensions like this definitely isn't going to help. Especially when the author is a self-proclaimed "not a developer" and made this by copy and pasting code from a LLM.

This new wave of non-developers deploying code to production that makes its way to a Show HN on the homepage here for 7 hours is scary to me. I'm all about making coding more accessible to non-engineers, but security, edge cases, failure states, etc is something LLMs are not particularly good at.

And even the non-code part of the original release of the extension had very little thought put into the block list. (E.g. he was blocking sites like the library of Congress, loc.gov, cancer.gov, CDC.gov, etc). Pastebin list of the originally blocked sites, which the author has since went back and cleaned up: https://pastebin.com/VHuHTeqp

> This new wave of non-developers deploying code to production that makes its way to a Show HN on the homepage here for 7 hours is scary to me.

I understand the feeling, and yet different tools/stages of increased digitalization of life are at the root of what is scary, not AI. personal data is input into the "global world computer" at astonishing rate, payments, payment credentials, all this attracting opportunist players with outright intentions of grabbing all this "public" data. even seasoned engineers are vulnerable to supply chain attacks. whether "new" code is developed by or with AI assistants seems the least of the problems in this context.

Both those things are stupid. Or scary. Whichever you prefer.

This isn't dev. This isn't hacking in the classic sense of the word. This is asking a genie to do it, trusting he's not malign or incompetent, and then saying "look, I made this!"

There are uBlock Origin filter lists to get rid of StackOverflow clones from web searches, such as:

* https://github.com/levymetal/filter-lists

(I don't remember where I took that one from, but probably was from this comment: https://news.ycombinator.com/item?id=33747333)

uBO already works fine on Chrome, Firefox desktop, and Firefox Android. Same syntax could be used to block any other domains, as desired. No need for a whole independent extension to just do that.

Completely OT, but I noticed in the last two days or so that I'm getting way more (often unskippable) ads before youtube videos, despite running uBlock Origin. Has anyone else noticed that too?
The word that's missing from the headline: "results"