I can weigh in on this. I run a small information website (https://allaboutberlin.com). My traffic has been growing steadily for 7 years, but now its down 30-ish percent year on year. From what I'm hearing, my competitors (and friends) are having it worse.

My main mission is to put new information on the internet. It's harder to do this if AI destroys the economics of it. It's also harder without an audience who provides feedback and encouragement. Having all information mediated by two companies isn't just digging into my revenue, it's killing the fun.

The unfortunate side effect is that it made me focus more on business and less on serving my readers. I'm working on health insurance stuff when I would rather work on a citizenship guide. It's also sad to see all the nuance stripped from my carefully chosen words.

It sucks, honestly. I am being replaced by AI, but I am still supposed to go experience the real world and translate it into LLM training material. In the end, there still is a thinking, feeling human being doing the essential work. He's just not getting paid anymore.

Thanks for sharing. I've been thinking of creating a non-revenue online guide to my new city, Lyon, France, but I've been concerned that I'd only be feeding the AI machine.

Ya, I'm in a similar boat (https://shepherd.com). Despite sky-high engagement stats, we lost 70% of our traffic from Google. Luckily, we have a strong brand, a growing community, and get decent traffic directly (and from other search engines).

We are surviving, but we've changed our entire focus as the web we know is a dead man walking. AI bots are the new frontend for the web, and I am curious about what the web will look like in 2 to 10 years. I hope the AI companies start paying to access the data as they answer queries or construct dynamic frontends (but I doubt it).

I have many friends who ran great indie websites, and it has been sad to see that 90% of them are now frozen or closed.

If you want readers more than impact you might want to consider something like a Substack. The paying audience would be smaller, but, it's not crawlable by AI (at least not today).

Still, it's a bit ironic that your own website advises people to use bots to get apartments (thus bypassing the work put into the adverts and apartment hunting sites), whilst also decrying bots who would bypass your site.

More generally perhaps there's a market for a Steam-like walled garden but for "old web" style sites. Crawlers would be controlled via technical means, micropayments would be built in, and use of AI would be strictly controlled. For instance maybe the garden uses AI to build link directories like DMOZ used to be, but the information itself isn't trained on so you're guaranteed to have organic readers.

A walled garden or paywalled website would go completely against the website's ethos. Everything is meant to be free for everyone. "Information wants to be free." Besides, those who stand to benefit the most from that information are the ones who can the least afford it.

The bots are not a solution, they're table stakes. Withholding that information would not do anyone any justice.

If your goal is for information to be as free as possible, then having AI read it is the best possible outcome. But it sounds like you have other goals too involving personal satisfaction, maybe community building etc. All very fair and reasonable goals to have, but they inherently conflict with "information wants to be free".

>Still, it's a bit ironic that your own website advises people to use bots to get apartments (thus bypassing the work put into the adverts and apartment hunting sites), whilst also decrying bots who would bypass your site.

I believe these are pretty different scenarios. There's an ongoing, massive shortage of apartments in Berlin, and bots here often aren't bypassing "work put into adverts and apartment hunting sites"—they're trying to get an application in at all before the listing is flooded with hundreds of applications and disabled. The content of the application has nothing to do with it.

(Of course, this just leads to an arms race of apartment-application-bots, but this is the situation the city's found itself in. Same goes for registration appointments, residence permit applications, and so on.)

If you have a zero sum game, any advice you give is creating an arms race. The only solution is to have fewer Berliners or more housing.

Or eliminate rent control, allowing prices to float to the level people are actually willing to pay. You only see bot-driven stuff like that in cases where desirable goods are being deliberately under-priced.

Just a little bit of encouragement: this is a really, really nice site that reminds me of growing up on the 1990's internet. You can tell you, "actually give a shit" about what you're sharing and that alone I think gives it credence and relevance. Don't let a silly thing like economics keep you from this worthwhile human achievement pal.

I came across your website organically and it's been a great resource for me in the last few months, including times when the answers presented by Gemini have been definitively and legally incorrect. Thank you for organizing all of this information so clearly.

> He's just not getting paid anymore.

Or the credit. Or even just the feedback.

Even LLMs get at least some feedback.

This is a surprisingly taxing change. There's a difference between giving free eggs to your neighbours, and giving free eggs to the Holiday Inn across town.

> My main mission is to put new information on the internet. It's harder to do this if AI destroys the economics of it. It's also harder without an audience who provides feedback and encouragement.

I also run an online website that provides valuable information. However, we have a very different philosophy. The mission is very not not to "put new information on the internet." Rather, it is to provide a valuable and much needed service for people in our niche. While we do greatly benefit from SEO, we also work very hard to cultivate our audience, both in terms of finding new people, and nurturing our relationship with the existing ones. "Information on the internet" has zero inherent value without an audience. If you're not working to build that audience (outside of third party platforms), then it's going to be very tough in the long run, and you'll run into situations exactly like this.

FWIW, our SEO traffic has remained shockingly steady, though the search console has reported a dramatic increase in "impressions," likely due to featuring our website in AI generated summaries. We'll see if things change over the long-term, but for now, we're not changing any strategies based on this.

The problem is that there is no audience without a third-party platform nowadays. At least not at a scale.

Google was the lesser evil, but I still fully expected to get killed by an algorithm update some day.

I vowed to ride this website into the ground if I have to. I believe in its value, even if I need a side job to support it.

There is no "nowadays." There never an era of big Internet audiences without third party platforms. The early web only had small audiences, and the reason it worked was that people were okay with that.

You could have a small audience of a few million users, not try to compete with the giants, and still, be serving millions of users (daily)

The internet was a beautiful place

A good strategy can involve betting on alternative distribution channels beyond search engines. Here are some ideas:

- Newsletters. You own your audience and your list is portable and interoperable.

- PWAs with push notifications. Be careful not to overwhelm users with too many messages. Make opt-out options easy to find.

- Door flyers with QR codes. A return to the basics that can still be effective.

- Social media. Still useful, but its reach and engagement seem to be declining.

I am providing information for specific situations. People don’t casually read about replacing lost keys until they lose their keys. It’s not entertainment.

Major free-rider problem, this is—one, however, that property rights could protect against: You own your data. If the OpenAIs of the world want to ingest your data, they pay you for it.

We need a new information backbone

That was the idea behind Internet2.

Never used it, maybe someone can explain how it has evolved.

This is a great information website, congrats, Google is happy: "thanks for feeding me", they will scrape all and display on their site in the nice table. Sorry about that.

My niche has been dominated by indian slop for a decade now who cares about a few robots getting in on the action?

Implicit in this article is the idea that you have to use Google, and that Google Search equals “the web”. I’ll be the boring nerd, I guess, who has to be the one to say there are many other search engines out there.

If you don’t like Google (I don’t), or don’t like where they are moving their products (I don’t) — please use a competing search engine.

I’ve used DuckDuckGo for years and I don’t miss anything about Google. There are plenty of other good search engines out there too.

It’s not particularly difficult to switch away and get an experience more to your liking, so it’s a bit baffling when Google Search product decisions are equated with burying the web alive. Just don’t use it!

You have control over what you use to search the web, but you don't have control over the incentives of publishers, the content of which google used to get big and monetize and is now cannibalizing.

That’s a fair assessment.

If the intent of the article was to say that Google is creating an environment which incentivizes publishers to lock away content from scrapers, then that’s probably correct in the medium and long term.

But I feel like the market described as being cannibalized here was already in the throes of death thanks to blogspam SEO. For many information categories, in my experience AI results are quite a bit better than the top-ranked results.

And typically for those types of results you want to quickly shuffle off general web search into something more specific like Wikipedia or a site with domain expertise.

> in my experience AI results are quite a bit better than the top-ranked results.

Only because the AI results haven't been infested with ads ... yet.

As soon as the VC lottery stops, the ads will get injected and the AI stuff will quickly go just as bad.

When "to google" has literally become a verb in several languages, you know it's not going to be that easy to convince literally hundreds of millions of people to use something different. For many unsavvy people, Google is and continues to be their window into the web, and they wouldn't even know they can have it another way.

I wish Duck Duck Go was fewer syllables so I could try to force it into conversations as a verb.

Why don’t you “kagi” that? Could work, but it’s not free (but very worth it IMO).

Definitely poor branding. It needs much better 2025 branding and it’ll probably take off in this AI search era where results are really poor tbh and very inaccurate at times.

let me quack that for you

Did you find that on quack?

They own duck.com. They should revert to that.

People said this about AOL for a long time, too.

I don’t think AOL ever topped 40 million users while a significant chunk of humanity uses Google. This entire conversation hinges on their unprecedented scale and reach, I don’t think this is an apt comparison.

People (including me) said this for a long time about Yahoo! too. Kinda ironic they both merged and basically died together…

Yawho!?

Jawohl!

Sure, and that also didn't vanish in a day. All I'm saying is that there is a ton of inertia, and people will not be switching to something different in an instant.

You miss the 'A' in AOL.

You don’t have to convince hundreds of millions of people to use something different, you can just use something different.

It's a seemingly patronizing joke meme, but I've repeatedly seen many of my relatives type in e.g. "Facebook.com" into Google search box, and then blindly click the first entry (which is completely at the mercy of being hijacked/prioritized based on either who's paying more money to Google, or latest fun extension / glitchy website they've installed).

I have not yet found a working approach to alter that behaviour - obligatory XKCD reference [0]

0: https://xkcd.com/763/

So true and the holy grail of being the first entry fuels the industry of reviews and stars which fuels the industry of bad review removal services etc and so on. This is such a vast ecosystem and so humiliating for the participants that it is only logical to let AI write, read and handle it. This is our chance to get a life. Lets not squander it.

DDG is perfectly serviceable in ways Google is not.

But - Google owns the household name game. Anyone who does not just use whatever is the browser default quickly switches to Google because they are used to it and conventional wisdom is they are the best.

FWIW in terms of results, Bing is fine, I just fucking hate all the extra shit on the pages. I just want plain text.

I've been using Brave search and it has had the AI answers for much longer than Google has. I would even go so far as to say Google stole the idea from Brave, but then someone will point out a different search engine that was doing it longer than Brave.

I used DuckDuckGo for 3 years, the only thing I found it useful for was the g! bang, in my opinion it wasn’t a good search engine.

What do you use now?

I switched to kagi like half a year ago and it does the job but iam still unsure if I want to pay the 20 bucks for that service.

I find that yearly works better for me psychologically for stuff like this.

I got a 1 yr professional of kagi just to try it. IMO the results work. I've never seen Google do better when I compare; I have seen the Google AI responses be consistently straight up wrong.

To me it's worth the cost knowing I'm paying a sustainable rate for a service. Plus I want no part in whatever the hell Google is doing these days with search.

I went back to Google but I’m pretty addicted to Claude with the search plugin. I think it’s insane.

I don’t like the Google AI summary thing because you can’t seem to “converse” with it.

kagi has an AI summary tool to which is pretty good as far as i can tell.

If Google Search gives you the experience you want, you should probably keep using it. I was discussing options for people who are dissatisfied with their product or product direction.

I find it extremely challenging to believe you actually think that. I can't even host my own email anymore without all of my messages getting filtered as spam, because if your email service doesn't play nice with Google, then you might as well not exist. Not to mention youtube, adsense, Android (Google is trying to kill AOSP to sink their claws more deeply into Android in light of the recent antitrust suits),

Google owns too much. They're easily on the same level as standard oil and the railroad tycoons.

When a company is paying millions annually, just towards psychologists, to keep people hooked, telling people "lol just turn the computer off" is disingenuous at best.

I've been using a private email server for decades now, but it's become clear that none of the email I send using it arrives at its destination. I can receive email with them, but not send.

If this was a sudden and complete change, you may have landed on a spam blacklist.

That's certainly possible. I've had repeated issues over the years with my compilers being identified as "malware" because the object code did not match the Microsoft C runtime library. I.e. the code was not a match for anything in their databases.

For one, I didn’t say anything about email or you had to boycott Google completely. You can just change your browser to use another engine it takes like 10 seconds max. Google Search is not the web.

Second, since you brought up email —- I’ve used Fastmail for a decade and it works just fine.

> Google owns too much. They're easily on the same level as standard oil and the railroad tycoons.

I agree 100% but since regulators largely refuse to deal with it, there’s nothing stopping individual people like you or me to make our own decisions. You literally just have to click a few buttons if you’re unhappy with the service Google Search is providing.

As someone who remembers what the web was like 25 years ago, I have a feeling that they're burying a corpse.

There’s still tons of personal webpages out there, Google just doesn’t expose them because it’s in their interest to show ad-filled SEO blogspam since they also operate AdSense and DoubleClick for Publishers. https://marginalia-search.com/ does a much better job of finding personal sites.

It doesn't really "find" any personal sites, you have to tell it about each manually with a pull request: https://github.com/MarginaliaSearch/submit-site-to-marginali...

This isn't true, Marginalia Search has a crawler. The guy who runs the site writes a lot about the trials and tribulations of running a web crawler on his blog if you're interested. https://www.marginalia.nu/log/ I just checked and my site is listed despite me never manually submitting it. I guess if nobody links to your site, you would have to manually submit it due to it not showing up in crawls, but that's true about any search engine.

That form is just to insert domains that aren't found automatically.

Is that website down, or maybe it doesn't work in EU?

Edit: NVM, just took a few refreshes to get past the "oh noes" screen

Works in NL

Right?

I'm sure Google deserves some, maybe a lot, of the blame, but the web has just deteriorated steadily over the past 15 years or so and I really don't think it's all their fault.

Wet streets cause rain-level observations from non-publishers in this thread. The web everyone is nostalgic about was killed by google.

The corpse was already buried ages ago and his name was Jeeves: What they are doing now is an exhumation and calling it the future.

Yes, a corpse killed by Facebook, Reddit, Discord and other closed platforms.

All the negativity is public and algorithmically boosted, all the positivity is in private chats. And this is increasingly the case over time. I sincerely hope for maximum enshittification in the form of AI slop rendering these misery machines discredited in the public's mind so the spell is broken.

Does it actually matter that Discord is closed or would the problem still exist if Matrix won and the content contained within was nonetheless still not indexable by search engines and communities were invite only?

Because the fact that Discord is flourishing is I think a direct result of it not being a part of the giant amalgamated omniplatform indexed by search engines. Being shielded from the masses is the point. There aren't megaphones in small private communities where the social dynamics of repeat interactions are in play. People aren't competing to say the most outrageous rage-bait for engagement.

But the counterpoint is that it's much harder to find any information, including personal anecdotes from people, without being part of the particular community.

>People aren't competing to say the most outrageous rage-bait for engagement.

This was never true in the age of web forums, which was the equivalent to Discord in many cases, and still isn't true in niche subreddits, which are public and indexed by search engines.

I'd like to limit my online time, but I'm so addicted to the internet, even though it's pure garbage nowadays. Any practical tips? It's not about putting away the phone, it's about... the hope that maybe this time I'll read something funny or informative, as I have absolutely no other sources of information or social interaction. "Go out and talk to people" isn't valid advice. It's never been, but in 2025 finally even the mainstream is slowly admitting that this doesn't work.

> "Go out and talk to people" isn't valid advice. It's never been, but in 2025 finally even the mainstream is slowly admitting that this doesn't work.

Why doesn't it work?

It absolutely does work but it’s too vague. A person in this position really needs to go outside and talk to people but doesn’t where to start.

So, join a running club, or go to pub trivia, take an in person class, or volunteer. Pick literally anything you can show up once a week and there will be people doing stuff. It will change your life.

Learn an ancient language and read nothing in English for 2-3 years. Once you come back to English, avoid anything after the 19th century.

This feels like the culmination of a long trend. Google shifting from indexing the web to replacing it. The idea of an “AI answer mode” burying actual sources is worrying, especially for niche or emerging topics where LLMs hallucinate confidently. I’d love to see metrics on how often users click through to original sources under the new interface. At some point, if Google becomes too self-referential, it risks losing the very web it was built on. Curious whether this opens the door for competitors that prioritise raw links and transparency over synthesised summaries.

I get why this is problematic for industries that depend on high traffic for ad revenue etc, but is bad for websites who are actually trying to market services that provide tangible value? Like if I’m searching for a dry cleaner in Glasgow, if I end up with the same provider, I don't care (and neither does the dry cleaner) whether I find them through traditional links or an AI-mediated search?

Feels like more curation than we had before. My wife and I recently bought a commercial space and have been trying to find contractors for renovation projects. It has kind of blown us away at how unresponsive most have been. Like, we'll reach out them ... wait DAYS, sometimes they get back to us ... maybe a week later they'll come for a site visit to give us a quote and then we need to keep following up with them ... just to get a quote.

We were complaining about this and a friend of ours said something very insightful: "There's a lot of good contractors out there who suck at SEO."

So we reached out in some local Facebook groups and had people out THAT DAY and the work was starting the very next day.

Point is, the companies we were reaching out to were the absolute busiest of the busiest because they were showing up at the top of search results. As soon as we stopped using web searches, we actually started getting results.

Now, this problem exists in spite of AI but I'm worried that it's going to make it even harder for smaller, less busy companies to show up AT ALL. AI is taking the system that already existed and is curating it further with "summaries" ... so what can we expect? Only the first "top 3" businesses in your area?

Google and other search engines were already starting to kill off the "long tail" ... but with AI I think its going to be the final nail in the coffin. If you want to search for the less popular options, if you want to try and find the niche stuff ... I don't know how well it's going to work. I suppose if you can prompt it: "All the 'top' options are ghosting me, give me some alternates" then maybe people will find ways around that with these tools. Or trying to find businesses with web search was already not going to give good results anyway in current year (return to classifieds and job boards was our recent experience).

Some "plumbing" company websites are really just referral farms that are supposed to send your data to plumbers in your area.

The idea even makes sense - if this worked, plumbers could focus on plumbing and SEO hackers could focus on getting a landing page in front of people looking for plumbing.

Sadly there are a lot of slop-like companies out there.

Or the AI takes care of finding quotes for you, in which case the most responsive vendors could do better than before.

>Like, we'll reach out them ... wait DAYS, sometimes they get back to us ... maybe a week later they'll come for a site visit to give us a quote and then we need to keep following up with them ... just to get a quote.

Jeesh these companies are dumb... they could be selling referrals to other businesses but are instead dropping them on the floor.

The best way to sell your boring contracting company is to take it off the web and wait for private equity to make an offer.

And if Amazon gets into dry cleaning and promoted Amazon Dry Cleaning as the top result in Glasgow: will that be bad for local Glasgow dry cleaners when everyone is shipping their clothes to Amazon?

Which is a problem for websites that only provide facts.

If your whole business model is getting organic search engine traffic to answer a user's question (a time, a result, something they just want to know) then Google is going to eat your dinner and I'm completely ambivalent because so many of the websites that have been so expertly sculpted to dominate SERPs, are what's wrong with the web. It's why it deserves to be buried.

Before posting this, I'd just seen this post about asking Forbes whether or not the latest Mission Impossible has a post-credit scene: https://bsky.app/profile/chrisplummer.bsky.social/post/3lpvx...

> If you load this page it contacts 82 IP addresses executing 256 separate HTTP transactions to download 18MB of data writing 64 cookies to your device to tell you “no”

The web has been on life-support for a while. Ad-blockers just make it bearable, but its existence as an information lookup system is nearly over. It's a way to interact with online systems and —occasionally— for reading some long-form content (which is increasingly written by LLMs anyway). Isn't the future confusing.

Isn't that a setting the individual publishing user controls?

I don't use BlueSky so I'm not sure - just interested. Especially if it's already going the route of Twitter/Facebook, etc. - which I expect it to long-term.

Yeah there's a "learn more" link.

> This user has requested that their content only be shown to signed-in users.

I used to hate the AI overview, but I've slowly been getting used to it and now I will often settle for what the AI gives. It's better than scrolling through loads of links and trying to weed out garbage SEO optimized sites.

At least in scientific writing, AI summaries have been proved to change the meaning of the results. So, use with caution.

How hard is it to get the infernal engine to work for you?

"Belphegor-2 go wade through the search results and discard the low quality machine slop, I only want results linking to curated human generated pages."

Results are OK, until you have this "garbage".

This is actually a good thing, now we don’t have to sift though a dozen semiautomatically created SEO gaming website that try and make the same point 5-6 said in slightly different ways.

I’m in two minds. On the one hand I want to consume quality original content. On the other hand when I ask a question, the top hit being the literal answer is such a convincing UX.

But then again, I have always found the quality content through social media, links to blog posts, YouTube channels, etc. I’m not putting “quality politics blogs” into a search engine and never have done.

What is better? A single source or a variety of sources to determine bias? If I do publish on the web? What is my incentive to keep doing so if my site never gets seen in search results? What will fuel the LLM then?

I haven't used AI Mode but AI Overview often reads exactly like those 300 word bullet point listicles if you try searching for anything competitive like "best X for Y"

So if Google innovates with AI, it's a problem because they're hiding their "ten blue links." But if they don't, they're suffering from the innovator's dilemma and are going to die with their ten blue links. The fact is, Google is so big that they're always going to upset some people.

Embracing AI and moving forward is the only thing they can do.

Why is it a foregone conclusion that embracing AI is the only way to move forward, and that embracing AI is necessary to succeed?

I'm certainly curious to see how it works out - it's a possible source of growth etc - but it's weird to me seeing people just assume that AI is not only The Future but is The Only Future.

If Google doesn't do this, people will slowly (or even quickly, who knows?) move to Perplexity, ChatGPT, etc. I know I personally would abandon Google entirely if it didn't summarize results and help me avoid the SEO slopfest.

Because AI/LLMs are clearly a 10x+ improvement over a significant number of use cases for search. And LLMs will be the thing performing the searches and filtering results in the remaining use cases.

It is because it is easier to imagine the current trend taking off and staying than imaging the alternative tech line that will beat both AI and the ten blue links.

I think publishers will not be pleased about a steep fall in click rates, now publishers still have considerable political influence. What will Google do in the event of serious legal pushback or a renewed drive for antitrust action?

The introduction of AI overviews into Google search will cost quite a lot in compute/other resources, despite heavy caching, therefore this might be a significant bet in terms of costs vs profit for Google. What does Google expect from this feature in terms of business results? This seems to be quite a big bet, but what is actually at stake - in real terms?

Come to think of it: is there now a showdown between Google and Microsoft/OpenAI, where collateral costs are no longer taken into account?

Story forget to mention that you can add "fucking ..." to your queries to eliminate the obnoxious and more often incorrect than not AI summaries.

My summary - G is 'burying' web search results underneath AI results on its search page.

Haven't they been doing that for at least a year?

And before that they were burying it under paid results and an info box?

Incomplete. The article also discusses the new AI search feature (not the AI box), which eschews links to the web completely

You fell for the Goomba Fallacy https://i.redd.it/33w3fqrnx8pd1.jpeg

Yeah yeah. But nonetheless group think and echo chambers are real things. Not all opinions are uncorrelated data points, probably most of them aren't. And real people (you too) do indeed hold contradictory opinions all the time.

In point of fact insisting that "My Opinion Is Pure And Special and Uncorrelated To Anyone Else's" is probably the single best way to find yourself parroting echo chamber delusions. It's a license to ignore everyone who disagrees with you, because you can always construct things such that they're really arguing with someone else.

I think it's very appropriate to point out when "General consensus on HN as measured by article upvotes" happens to include two grossly contradictory opinions.

> Yeah yeah. But nonetheless group think

That's only your perception of groupthink. You read those threats looking for anti-Google bias, and you find it, because of course.

AI is a very divisive subject nowadays. There are people that are all in on the hype, there are skeptics, there are desperate people thinking that they will lose their jobs, there are those that think is is destructive and unethical.

The only caveat that I might make is that at the time of the Microsoft announcement, it was during the honeymoon phase of OpenAI (when GPT 3 was fairly new and an impressive leap from what came before), and Google's current advancements, while impressive, come at a time when a lot more people are negative towards it.

Also, every sufficiently large company will have people that dislike it as a baseline. Be it Google, Microsoft, Nintendo, Sony, Tesla, Toyota or any other. Read any thread on those companies and you will find negative comments based on that alone.

> That's only your perception of groupthink. You read those threats looking for anti-Google bias, and you find it, because of course.

Or... maybe it's only your perception. You're engaged in groupthink and read headlines looking for Google crimes[1], and you find them, because of course. I mean, I'm the one citing evidence (literally the headline we're discussing). Your theory that we're actually seeing a general change in attitudes toward AI in the HN commenter community seems interesting! But... unsupported.

Occam says you're just engaged in apologia for the echo chamber, sorry.

[1] I mean, come on. "Burying the web alive" really seems like reasoned discourse to you? A real rationalist would enter this discussion with a prior that the ridiculous hyperbole is probably wrong, not "let's find ways to make the ridiculous hyperbole more reasonable".

> I mean, come on. "Burying the web alive" really seems like reasoned discourse to you?

No, I actually disagree with the rationale. If anything, AI makes it easier to find some quick answers, since Google's top results tend to be shit.

I am very ambivalent towards the current AI hype. I think that the capabilities of LLMs are vastly overrated, even if they are useful to an extent.

That said, this it not to say that I don't think Google was very damaging to the web in many ways, mainly with how it made search so much worse over the years.

> A real rationalist would enter this discussion with a prior that the ridiculous hyperbole is probably wrong

Eh, I can understand that with the more pronounced negativity towards AI nowadays, this sort of rant will become more common. Every thread on AI nowadays has this overall vibe (not without reason).

> Occam says you're just engaged in apologia for the echo chamber, sorry.

Eh, I actually don't think very highly of HN. Plenty of terrible takes on every thread.

This is just a glorified Reddit full of people who really like the smell of their own farts. Why would I defend it?

So how long do we have until companies can pay to be in the system prompt to be recommended for certain queries?

My speculation: It's almost certainly already happening.

Yes, it's called "tool partnerships" and Microsoft has been doing it heavily for Copilot.

Can’t you just game the SEO and Google will eat your result thus pointing you to whichever service/product?

It’d be spam generated answers from seo spam. Which is definitely already happening.

Those AIs are operating at a loss. Sooner or later they will seize enough market share for investors to demand returns.

How to fix Google:

1. Install User-Agent Switcher for Chrome plugin.

2. Choose older browser, like IE 10 or Iphone 6, refresh the page.

3. You're now getting "old google" results. Seeing first 5 links without scrolling. Accurate and concise.

The contrast is STARK. There is no "AI overview". There is no "From sources across the web". There is no "People also ask". No "People also search for". No "At a glance" section. No "Images" section, nor "youtube" section taking 50% of the page. No "Short videos" tab. Just. Search. Results.

SEO as we knew it gone forever, Google push with its AI drove everything to shit.

I used to use Germini and it was pretty good, then it went downhill with wrong answers and I just couldn't trust it anymore.

That is where I crossed Perplexity ( https://www.perplexity.ai/ ), it is surprisingly solid and the best part of it: the sources. It even helped to learn Python.

It does mention all the sources it used for that answer, so in doubt, you can cross check which is something Gemini failed miserably to do so. There is very minimal "wrong answers" and in doubt, cross check it.

I went De-Google years ago with GrapheneOS and everything, I still use Google dot com to check stores reviews because unfortunately, no other search engine provides that info. There are websites only for that but it is not remotely close to Google reviews left by folks.

Computer owners finance www through Internet Service Provider subscriptions fees. Users of www upload content at their own expense. Google produces zero content at zero expense, stores content of www users, monitors their www use. Google profits, becomes trillion dollar company. What happens next.

What I want from Google is showing me results (not summaries) from sources that I tell it from natural language. If I know I am going to search for a topic I know is going to be filled with a ton of SEO garbage, I default to chatgpt to get sources for me, unfortunately I still do not trust chatgpt to have done a thorough job but sometimes I can't be bother to trawl through blogspam.

The reason why chatgpt and perplextty are stealing queries from them is because Search has become so bad.

There’s no web. There’s big ad companies striving to keep eyeballs. There’s a few lone bastions of what was. Mostly it’s ads.

I'm finding that google's own gemini flash is cannibalizing google search in my own usage, but there's no other way for them to survive being eaten alive by the competition.

The sad realization is you probably shouldn't expect humans to read your content anymore - it'll be quoted verbatim (best case) by the all-consuming LLMs, just without your ads, your layout or anything yours at all that isn't your written or spoken words. (Sad part is that's often an improvement until chat interfaces start putting product placement in their responses. We can truly bury the web then.)

Perhaps we need legislation requiring LLMs to compensate the sources they steal from? If the LLMs kill the hand that feeds them, it's bad for everyone.

It may reduce the tendency to visit websites. But I do the same with YouTube videos. I share the link with Gemini and summarize it. I think Google will not allow this after a while.

I only do physical offline marketing for some time now. Works well enough and one stands out among all the clickbait and other foolery. Some index disappearing is not a problem. The moment organizations or institutions start burning the libraries down, then we have a bigger problem.

Don't get fooled by people who claim the open web is dead. The moment you start believing that, then you are exactly where they want you to be.

I have to resort to using ChatGPT to find any historical information because of the recency bias in all conventional search engines. Never mind their penchant for stuffing results with e-commerce sites no matter how irrelevant they are to the query.

Yeah when I'm scrolling past the AI overview to click on the links it feels like I'm digging out someone buried alive with my bare hands to breathe new life into him.

Business idea: build a firewall for AI crawlers that variably gives them access to customer’s sites based on traffic they send. The more traffic they send, the more they are allowed to crawl.

Cloudflare has been talking about this for a while albeit slightly different than your take - https://blog.cloudflare.com/ai-labyrinth/

Yeah that was actually what I was thinking of for the idea, but the problem with their firewall is that it doesn’t penalize the big crawlers, like Google’s bot, because Google still sends traffic. Even if Google is sending less and less traffic over time.

That just gives the biggest companies the prize of being able to crawl all the web for free, while they offer less over time in exchange.

The web has been a walking corpse for about a decade now.

The hardest issue in "fixing" this is "reminding people that the web is open and you can kind of do whatever you want here."

I've said this before, but I sincerely believe that a human-curated internet directory would be a great resource. Maybe something along the lines of Yahoo Directory revived? Organised using the dewey system could be useful. A library for the internet.

I'd love to be able to look up something like say 'shoe repair' and find a host of high quality resources such as videos, tutorials, suppliers, etc, all vetted by real people somehow.

Surely, someone must have already thought of this. Perhaps what we really need is a curated index of curated indices.

I think the answer lies in a network of curated directories, and search within a localised range of the network.

Many people already have a decent list of curated sites in the form of bookmarks, but bookmark software is horribly dated. Even if I limited my search to sites gathered from my own bookmarks, I'd find 95% of everything I'm ever searching for.

Forgive this weirdly philosophical comment: In a year of dark thoughts I found yet another horror to contemplate, after reading a legal paper on the idea of the marketplace of ideas.

A core concept underpinning our shared market place, is that its not about facts, its about ideas being exchanged.

Critically, that exchange creates friction, debate and downstream of all that - society. I've commented elsewhere how the marketplace is broken today, and why.

But with GenAI, theres a new, twist, that I can barely contemplate.

See, our conception of this liberal democracy based society, exists in that exchange between people. Even a crappy exchange, is part of the process of learning and building our ideas. My half formed ideas, your half formed ideas, everyones ideas.

We have considered what happens when your ideas don't get heard.

But in no place, did we consider what happens when we bounce our ideas into a machine, and that machine talks back. And if that machine talks back in a manner thats can be as good as what a person might write or text. Your half baked idea goes into the machine, you get text back that helps you get to a fully baked idea.

But the counter party is the machine. Not a person. The way those ideas have been bouncing around, has changed.

I noticed that Google image reverse search turned into completely unusable garbage that can only produce AI lookalikes but completely fails at finding original images. Literally every time, exact match - zero results.

I guess that's their current trend. Forget about quality search, inject AI everywhere.

AI summaries aren't the problem, they are a poor solution to a overwhelming amount of slop content. Search engine results have been getting worse and worse for over a decade, and with a rapidly increasing collection of content the web is getting too big to search. AI will be necessary, but it's not going to change the course, it will be deployed on both the solution and the problem front.

If I had to pick the biggest contributor to the problem it would be advertising. Which mainly means Google, and governments for allowing an ad and search monopoly under the same roof. People can make all the justifications they want about how ad revenue allowed businesses and individuals to prosper, and "free" access to content, but it was always borrowing against the future.

Maybe Google should use AI to de-rank ad filled seo spam.

the only thing google will win doing this in the long term is new laws that you'll need to opt in for training AI models

Correction: burying our business model alive

In my humble opinion, Google is the reason the web withstood as long as it did. At least the open part of the web did.

Yes, sure, they monetized, but also they gave back as much as, and if not more than they took. We have so many machine learning frameworks, tensorflow, research, payouts to creators, advertising opportunities, careers, products, a lot of things built and taken down but most importantly built. They were probably the most positive force for the internet age in the past 20 years and more than anyone will ever give them credit for. Only in retrospect will we realize how lucky we were to be alive in the Google age. Full stop.

What really killed the web was the rise of closed wall gardens platforms such as Apple, Facebook, Instagram and others. Putting up walls around content that didn't need to necessarily exist or not honoring open frameworks to exchange information and making things more widely indexable.

But even here there have been significant benefits. The present AI boom would arguably not have been as large as it is right now without Mark Zuckerberg choosing to put an unconventional amount of investment behind AR ambitions to take on Apple, an investment the size of which many conventionally run publicly listed or private enterprises could hardly imagine to take up, leading to the concentration of capital, talent, technology and hardware in a place that gave birth to open source Llama and others. Google as well was very well poised because of their investments in compute fueled by their business model which kept the web alive and also returned capital into places where computer scientists would be paid significant amounts of money and have job security and freedom of will. to do as they pleased as opposed to chasing a paycheck, working as a physicist at CERN or something.

All I'm saying is this article does not fully capture how significant the positive outcomes from Google have been.

it's so funny listening to people who are clearly not publishers give their takes on these things

Google has been getting this push back for the last 15 years. (That makes me feel old to remember it)

Google Knowledge Graph (that sidebar they show) was hated by publishers, especially Wiki for stealing their content.

Google adding direct answers to questions.

Lots of fights over social media & recipe results.

I'm not arguing who is right or wrong,* just saying this has been a thing for a long time.

* Exception most recipe websites & those infinite looping Pinterest blog links. Those websites are all awful & wrong.

Surely as the traffic Google sends to websites falls and falls, there'll come a tipping point where the websites themselves will feel there's no benefit to allowing Google to index them.

If this happens, Google's generative AI slop will have less and less scaped and regurgitated content to surface to Google users.

Lazy publishers could have solved this years ago, by making massive syndicates, paid by a single subscription. YouTube did it, and has become a modern wonder of the world.

But no, of course they can't syndicate together with other newspapers or magazines who aren't 100% psychotically equally aligned with them politically.

So they get what's coming to them. They shouldn't have been megalomaniacs, but that's all they are now.

Soon, quick and dirty solution: User-Agent: * Disallow: /

So then the internet police know to zoom up the fiber line and arrest the user-agent for not following robots.txt!

Goodbye small blogs and medium publishers, it was nice to read you. Now Google take your unique content, rewrite and present on their website. There will be no incentive anymore to write a new content. You will get NOTHING for your hard work. Bye.

You can't simultaneously rail against ads, which is our best model for scaled monetization, or complain against paywall and then say it is killing the business model.

Stuff costs money but unfortunately we as a society have come to expect free stuff, and businesses are just catering to it.

If not for Google, ChatGPT or someone else will do it.

literal mind control

Google is damned-if-they-do, damned-if-they-don’t. One side screams that Google Search is dead because Google doesn't innovate. The other - Google AI innovation is killing the web.

I find it very hard to believe that any of the major platforms is going to defend against AI slop. They're looking to monopolize it, which will reduce but not stop it in the short term, then maximize it later on for maximum dollars.