back
845 comments
The overviews are also wrong and difficult to get fixed.

Google AI has been listing incorrect internal extensions causing departments to field calls for people trying to reach unrelated divisions and services, listing times and dates of events that don't exist at our addresses that people are showing up to, and generally misdirecting and misguiding people who really need correct information from a truth source like our websites.

We have to track each and every one of these problems down, investigate and evaluate whether we can reproduce them, give them a "thumbs down" to then be able to submit "feedback", with no assurance it will be fixed in a timely manner and no obvious way to opt ourselves out of it entirely. For something beyond our consent and control.

It's worse than when Google and Yelp would create unofficial business profiles on your behalf and then held them hostage until you registered with their services to change them.

In the UK we've got amazing National Health Service informational websites[1], and regional variations of those [2]. For some issues, you might get different advice in the Scottish one than the UK-wide one. So, if you've gone into labour somewhere in the remote Highlands and Islands, you'll get different advice than if you lived in Central London, where there's a delivery room within a 30 minute drive.

Google's AI overview not only ignores this geographic detail, it ignores the high-quality NHS care delivery websites, and presents you with stuff from US sites like Mayo Clinic. Mayo Clinic is a great resource, if you live in the USA, but US medical advice is wildly different to the UK.

[1] https://www.nhs.uk [2] https://www.nhsinform.scot

I still find it amazing that the world's largest search engine, which so many use as an oracle, is so happy to put wrong information at the top of its page. My examples recently -

- Looking up a hint for the casino room in the game "Blue Prince", the AI summary gave me details of the card games on offer at the "Blue Prince Casino" in the next suburb over. There is no casino there.

- Looking up workers rights during a discussion of something to do with management, it directly contradicted the legislation and official government guidance.

I can't imagine how frustrating it must be for business-owners, or those providing information services to find that their traffic is intercepted and their potential visitors treated to an inaccurate version on the search page.

I was at an event where someone was arguing there wasn't an entry fee because chatgpt said it was free (with a screenshot of proof) then asked why they weren't honoring their online price.
Fun. I have people asking ChatGPT support question about my SaaS app, getting made up answers, and then cancelling because we can’t do something that we can. Can’t make this crap up. How do I teach Chat GPT every feature of a random SaaS app?
I particularly hate when the AI overview is directly contradicted by the first few search results.
Google AI has been listing incorrect internal extensions causing departments to field calls for people trying to reach unrelated divisions and services, listing times and dates of events that don't exist at our addresses that people are showing up to, and generally misdirecting and misguiding people who really need correct information from a truth source like our websites.

Anecdotally, this happened back in analog days, too.

When I worked in local TV, people would call and scream at us if the show they wanted to see was incorrectly listed in the TV Guide.

Screamers: "It's in the TV Guide!"

Me (like a million times): "We decide what goes on the air, not the TV Guide."

This raises the question of when it becomes harmful. At what point would your company issue a cease-and-desist letter to Google?

The liability question also extends to defamation. Google is no longer just an arbiter of information. They create information themselves. They cannot simply rely on a 'platform provider' defence anymore.

Their goal has always been to be the gatekeeper.
Self replying for context so I don't read this a few years and question whether I'd missed a sign of early-onset dementia.

You're fine, you just lost a few verbal IQ points after fasting for 24 hours and doing blood work.

Of course slow, shitty web sites also cause a massive drop in clicks, as soon as an alternative to clicking emerges. It's just like on HN, if I see an interesting title and want to know what the article is about, I can wince and click the article link, but it's much faster and easier to click the HN comments link and infer the info I want from the comments. That difference is almost entirely from the crappy overdesign of almost every web site, vs. HN's speedy text-only format.
Here is the experience when clicking a link on mobile:

* Page loads, immediately when I start scrolling and reading a popup trying to get tracking consent

* If I am lucky, there is a "necessary only". When unlucky I need to click "manage options" and first see how to reject all tracking

* There is a sticky banner on top/bottom taking 20-30% of my screen upselling me a subscription or asking me to install their app. Upon pressing the tiny X in the corner it takes 1-2 seconds to close or multiple presses as I am either missing the x or because there is a network roundtrip

* I scroll down a screen and get a popup overlay asking me to signup for their service or newsleter, again messing with the x to close

* video or other flashy adds in the content keep bugging me

This is btw. usually all before I even established if the content is what I was looking for, or is at any way useful to me (often it is not).

If you use AI or Kagi summarizr, you get ad-free, well-formatted content without any annoyance.

I've written high-quality technical how-tos for many years, starting with PC World magazine articles (supported by ads), a book that helped people learn Ruby on Rails (sales via Amazon), and more recently a website that's good for queries like "uninstall Homebrew" or "xcode command line tools" (sponsored by a carefully chosen advertiser). With both a (small) financial incentive and the intrinsic satisfaction of doing good work that people appreciate, I know I've helped a LOT of people over four decades.

A year ago my ad-supported website had 100,000 monthly active users. Now, like the article says, traffic is down 40% thanks to Google AI Overview zero clicks. There's loss of revenue, yes, but apart from that, I'm wondering how people can find my work, if I produce more? They seldom click through on the "source" attributes, if any.

I wonder, am I standing at the gates of hell in a line that includes Tower Records and Blockbuster? Arguably because I'm among those that built this dystopia with ever-so-helpful technical content.

A lot of the comments here are along the lines of "websites are often hostile, and AI summaries are a better user experience" which I agree with for most cases. I think the main thing to be worried about is that this model is undermining the fundamental economic model the internet's currently based on.

If I create content like recipes, journalism etc, previously I had exclusive rights to my created content and could monetise it however I wanted. This has mostly led to what we have today, some high quality content, lots of low quality content, mostly monetised through user hostile ads.

Previously, if I wanted to take a recipe from "strawberry-recipes.cool" and published it on my own website with a better user experience, that wouldn't have been allowed because of copyright rules. I still can't do that, but Google can if it's done through the mechanism of AI summaries.

I think the worst case scenario is that people stop publishing content on the web altogether. The most likely one is that search/summary engines eat up money that previously came from content creators. The best one is that we find some alternative, third way, for creators to monotise content while maintaining discoverability.

I'm not sure what will happen, and I'm not denying the usefulness of AI summaries, but it feels easy to miss that, at their core, they're a fundamental reworking of the current economics of the internet.

At some stage Google will need to be accountable for answers they are hosting on their own site. The argument of "we're only indexing info on other sites" changes when you are building a tool to generate content and hosting that content on your own domain.

I'm guilty of not clicking when I'm satisfied with the AI answer. I know it can be wrong. I've seen it be wrong multiple times. But it's right at the top and tells me what I suspected when I did the search. The way they position the AI overview is right in your face.

I would prefer the "AI overview" to be replaced with something that helps me better search rather than giving me the answer directly.

I subscribe to Kagi. It's been worth it to have no ads and the ability to uprank/downrank sites.

And there's no AI garbage sitting in the top of the engine.

Conversely, it's useful to get an immediate answer sometimes

6 months ago, "what temp is pork safe at?" was a few clicks, long SEO optimised blog post answers and usually all in F not C ... despite Google knowing location ... I used it as an example at the time of 'how hard can this be?'

First sentance of Google AI response right now: "Pork is safe to eat when cooked to an internal temperature of 145°F (63°C)"

You can apparently disable these annoying and useless "AI" overviews by cursing in the query:

https://arstechnica.com/google/2025/01/just-give-me-the-fing...

Why is this being framed as a problem? People are obviously happier with the new feature, duh

Of course they need to make the AI overviews suck less, but saying it’s unfair to sites is crazy talk because your site now just generates less value than an AI response if that’s what stopped you from going

If you have content better than Gemini I will still go to your site

The tricky thing for Google will be to do this and not kill their cash cow ad business.
Related, but to whichever PM put the "AI Mode" on the far left side of the toolbar, thus breaking the muscle memory from clicking "All" to get back from "Images", I expect some thanks for unintentionally boosting your CTR metrics.
What I don't understand is how do you square that with google still reporting growth in ads revenues. Surely google's business should tank by that much (not even taking into account the additional cost of processing a seach with AI).
Main reason:

Google seach ranking already became so bad that finding the right query to produce a decent result became a craft in and of itself.

HN readers might not realize this, but I used to constantly have to search things for friends and family as they just couldn't find it with Google.

The AI result is a godsend fir them. I saw a massive drop in their requests to me.

The clicks in question: "Here's a thirty page story of how grandma discovered this recipe... BTW you need to subscribe/make an account/pay to view the rest of the article!"
If Google just copied and pasted the content from my blog and put it on their own website which had ads that paid them, preventing customers from going to my blog (and thus seeing my ads which I get paid for), that would be obviously wrong.

How is this any different?

Liberating me from "search clicks" is not a bad thing at all. I suspect many of us though don't even go to <search engine> anyway but ask an LLM directly.
My site never got much traffic from Google and I never wasted time on SEO. So I find it hard to care about losing clicks I didn't have.

But I do find it concerning that Google is effectively stealing the time I spend on research and not offering proper credit. I'm always careful to credit people and provide extensive links to the work of others on my site. But here Google and others are simply stealing that work, getting it wrong, and then claiming they did all the effort.

A lot of startups like Profound are going after the "AEO" (basically SEO but of AI engine) space and absolutely blowing up. Marketers are freaking out. It's crazy to see the search juggernaut finally getting threatened.
For some (unknown to me) reason the web pages that used to sustain themselves on very small, quiet ads that were shown on sides of the page, now require full page, annoying, shouting, focus stealing, one-minimal-x-closing-after-few-seconds ADS. Therefore AI summary is a panaceum for this sickness.

However I understand that pages that have articles written need money to pay for those articles. For some another (unknown to me) reason reading something by AI and summarizing it is not a copyright violation, but the work that has been done to prepare the valuable content (article) will not be rewarded, therefore extrapolating it to future I can see that the 'dead internet theory' is coming right up.

People will stop writing because they don't get rewarded, so AIs will start speaking and discussing with other AIs and the entropy might lower (as in the amount of information present in the internet) leaving us to AI generated content only.

Which is of course Google’s short-sighted goal. See also their push to switch to full “AI mode” search which doesn’t show results at all.
In a sense this is similar to what Amazon has been doing in few countries. Find top selling products, get them cheaper from somewhere, rebrand them, rank them higher and sell them. They don't need to invest in market research like their competetors, they get all data from Amazon.com

At big tech scale, this is clearly anti-compete and piracy IMHO.

Maybe if the search wasnt full of ads and scams
people forget why users search - to find what they are looking for. as the saying goes, “No one wants a drill, they want a quarter inch hole.”

the first time i realized Google had a problem was when i used ChatGPT to search for Youtube videos, and compared to Youtube’s search, it was an order of magnitude easier to find the exact videos i was looking for.

Hallucinations are not a problem in a query like this, because i have what i need to evaluate the results: did i find interesting Youtube videos to watch? did i find what i was looking for?

generally speaking, users seek to minimize the effort required to achieve their goals.

There is (IMO) still a lot of value in having a searchable Internet index. If AI summaries were the only option it would be very difficult to know whether something is influencing the AI.

The problem is, of course, that we've already lost that battle with Google and other search providers. We know they're influencing the results. There is no open Internet anymore, but at least we can check the different search indices against each other. Checking different AI summaries against each other seems a pretty fruitless endeavor.

Google introduced its Google Knowledge Graph in 2012 and for sure it caused the drop in website clicks but I would say not that significant. People want quality content and they will always look for quality websites with or without AI overviews. My digital marketing professor once said that customers who look beyond first page results are the most valuable ones, because they are dedicated users and customers. People who read AI overviews won't be your first class users and customers.
What is the long term plan for data acquisition?

1. Use existing websites for training data

2. Replace search traffic with AI prompts, thereby destroying the economic incentive for websites to publish data

3. ?

Not any AI tho.

I have replaced SEO with Perplexity AI only. It isn't a chatbot but it actually search for what you are looking for and most importantly, it shows all the sources it used.

Depending on the question I can get anywhere from 10 to 40 sources. No other AI service provides that, they use the data from their training model only which in my experience, is full of errors, incomplete, cannot answer or altogether.

Why could i perform a web search that leads me to multiple results with cloudflare identity checks and ad spam when i can just use a llm..
Knowing that your content is unlikely to be seen by many humans, and is mainly just feeding an AI is pretty discouraging. AI is great at detecting patterns, if content creations is discouraged, and more AI content is generated, AI content is likely to become so homogeneous that the internet becomes more of an echo chamber (than it already is).
I think it's a google problem. The AI summary is often times wrong so I don't bother. But the pages are garbage as well, so normal users won't click on that looking for information. Dark forest and all that. Pages started to optimize for Google and not humans. Humans don't visit anymore. Duh
To be fair, Google's actual search couldn't be much worse than it was lately. It's like they really try to get all the spam, clickbait and scams right at the top.

The AI overview sucks but it can't really be a lot worse than that :)

What current incentive is left for people to publish on the internet?

Discovery mechanisms are less effective than before, people no longer visit your website, copyright is dead and in the end we're just feeding an insatiable machine with slave labour. Or am I missing something ?

I've written a bit about this issue as well, coming from someone working at a company that partially relies on organic search results. It's been interesting (And slightly worrying) to see the shift and layoffs happening here already.

https://sasjakoning.com/blog/how-google-s-new-ai-overview-is...

I have to say that the comments here do offer a great alternative perspective as to how terrible the web has been to navigate before AI Overviews became a thing.

And thank God! It does. I’m done dealing with bazillion shitty websites with bad, slow, performance, bad ui and dark patterns. All I need is an information in a convenient format and that is what AI tools provide to me.
For the most part there really is no need to use search in the traditional sense for knowledge. For information it’s still the only because llms are not reliable. But ChatGPT must have taken a huge dent in google’s traffic.
I hope the small web will thrive again once the profit incentive for putting content on the web ceases to exist so SEO/dark-pattern heavy players will give up and stop suppressing valuable (altruistic?) stuff.
I feel like there should be two modes - search and answer. It's helpful to get an AI summarize answer like 90% of the time. But the other 10%, I am searching for something specific.

I use Perplexity's Comet browser part-time now. I feel like what makes me annoyed about AI overview summaries is that sometimes I really just need to get the actual site itself and not a summary. For example, I needed to find EC2 pricing information and Perplexity's default search gave me a summary answer. I was looking for Vantage.sh's instance info site.

> Google users are less likely to click on a link when they encounter search pages with AI summaries

Well, doh:

"Do I just read the AI summary? Or click past five pages of ads and spam to maybe find an organic link to something real?"

This means searches are still happening, just being routed elsewhere?

I noticed Google's new AI summary let's me click on a link in the summary and the links are posted to the right.

Those clicks are available, might not be discovered yet, curious though if those show up anywhere as data.

Google being able to create summaries off actual web search results will be an interesting take compared to other models trying to get the same done without similar search results at their disposal.

The new search engine could be google doing the search and compiling the results for us how we do manually.