Why Google’s New AI-Saturated Search Page Will Be A Disaster
from the the-end-of-ten-blue-links dept
Google didn’t invent full-text search of the Internet – that honor belongs to early pioneers such as WebCrawler, Lycos and AltaVista. But for the last 25 years or so, Google has been synonymous with online searching, providing the quickest and most effective way to find things online (although its results may be getting worse.) More recently, it has been adding to its search engine more features based on generative AI, first with its AI Overviews in 2024, and then a year later with its AI Mode in Search. Now it has announced the latest stage in that evolution with what it calls “A new era for AI Search”:
It’s more intuitive than ever, dynamically expanding to give you space to describe exactly what you need. Designed to anticipate your intent, it also helps you formulate your question with AI-powered suggestions that go beyond autocomplete. And you can search across modalities, using text, images, files, videos or Chrome tabs as inputs.
This new incarnation effectively turns search into a chatbot:
You can easily ask a follow-up question right from an AI Overview, and flow into a conversational back and forth with AI Mode. Your context stays with you, and as you explore more deeply, the links and supporting articles get even more relevant. This seamless experience is live today across desktop and mobile, worldwide.
As the the screenshot of the new interface above shows, the traditional search result links that are currently placed under the AI Overview have now been confined to a small panel on the right-hand side of the screen, which shows a cut-down version of today’s list. Users are encouraged to ask follow-up questions from the AI search chatbot, rather than exploring the links themselves.
What this is likely to mean in practice is that even fewer people will follow links to sites, something that was already happening last year; instead, they will engage with Google’s chatbot to gather information indirectly. This is terrible news for access to knowledge because it frames the Google AI search engine as the fount of all knowledge – one that will do all the hard work of finding information and combining it into an easily digested answer that can be interrogated further. It can do that because it has already ingested billions of Web pages and other information sources as part of the Large Language Model (LLM) training process. But search engine users will no longer know what some of those sources are unless they painstakingly click on the links in the new panel.
Most people will not bother, because the AI-generated results will be good enough – or at least will appear to be good enough. Unless visitors to the site take the trouble to follow the links to the sources they won’t really know how reliable those results are. For example, it is possible that the sources are wrong, or misleading; moreover, Google’s LLM may itself introduce new errors and distortions. There is also the question of how Google will insert ads into this AI-generated information, and to what extent advertisers will be able to buy preferential treatment in results.
This new mediated approach is clearly terrible news for Wikipedia – an issue already discussed on Walled Culture earlier this year – and for creators. Google will use the information found in their works, but will not actively encourage people to visit the originals. For many people, summaries will be good enough, and they will never discover the greater riches of the sites and creations that Google’s LLM is based on. Worse still, the original creators such as Wikipedia may not even be mentioned in answers that involve aggregating information from a large number of sources.
Similarly, the new Google search is the publishing industry’s worst nightmare. Not only is Google drawing on material they have published, but it is pushing links to those sources into the background. It seems inevitable that the Web traffic to publishers will fall yet further, making already struggling business models based on advertising even more precarious. That will have knock-on consequences for the funding of many sites – particularly newspapers and magazines – and for the commissioning of work from journalists and other creative professionals. Users won’t even need to visit Google Search much in order to keep up-to-date with topics of interest thanks to Google Search’s new agentic capabilities that will do the work for them in advance:
With information agents, you can stay updated on whatever matters most to you. Your agent will intelligently look across everything on the web, like blogs, news sites and social posts, plus our freshest data, such as real-time info on finance, shopping and sports, to monitor for changes related to your specific question.
In this case, not only will people not visit sites, but the latter will be constantly bombarded by various AI bots seeking information on behalf of users – increasing site running costs, and making sites less usable by humans. Another key announcement from Google will lead to a further flood of agentic activities that will pose new challenges to businesses:
We’re also expanding agentic booking capabilities in Search to a wide range of new tasks, including local experiences and services. Just share your specific criteria — like finding a private karaoke room for six on a Friday night that serves food late — and Search brings together the latest pricing and availability with direct links to finish booking through the provider of your choice. And for select categories like home repair, beauty or pet care, you can ask Google to call businesses on your behalf.
What emerges from Google’s latest announcements is less of a search engine, and more of an immersive virtual environment that is designed to keep people engaging with Google’s services, asking them for information, advice and even delegating actions to them. There is no doubt that many users will find these new features attractive, not least because they can use “conversational voice features” in Gmail, Docs and elsewhere. These are the digital assistants that have been promised for many years, able to understand spoken commands, provide information verbally, and carry out complex operations on behalf of users without the need for any complex training. For many people, that will be a boon, and they will doubtless migrate from the traditional search page, which will still be the default – at least for now – to the latest AI-infused version.
But these impressive technical features come at a high price, even leaving aside issues such as the environmental impact of the huge server farms they require. With the latest incarnation of its search engine, Google is making the World Wide Web as we have known it for over 30 years invisible, and therefore increasingly irrelevant to most people, who will be happy to let Google become their universal user interface to everything. And yet Google still depends on the Internet to supply all the information it is analyzing and repackaging. It risks killing the very thing that sustains it.
There’s another, more subtle issue. The new Google search features make finding information and carrying out actions very easy in many ways. Leaving aside the problem that this will require people to trust what is in effect a huge black box, where the internal workings cannot be examined, with all the loss of control this implies, there is another danger. People who use Google’s powerful new AI search services to offload many of their day-to-day actions may gradually lose the ability to understand the world and to act within it without that constant help. Such a dependence may be great for Google and its advertisers, but it surely cannot be a good thing for the future of society.
Follow me @glynmoody on Mastodon and on Bluesky. Originally published to WalledCulture.
Comments on “Why Google’s New AI-Saturated Search Page Will Be A Disaster”
“It can do that because it has already ingested billions of Web pages and other information sources as part of the Large Language Model (LLM) training process.”
It can do that because it has already stolen billions of web pages of copyrighted content.
There is a fundamental and obvious difference between crawling web pages to build a search index (fair use) and crawling web pages to incorporate them wholly in an LLM (a clear copyright violation). Google, like the other AI/LLM companies, has engaged in massive and deliberate copyright violation, vastly in excess of anything ever done by all music/video pirates, and it should be punished with 12-figure fines and lengthy prison sentences for all personnel involved.
Re:
“They stole muh words!”
Disaster in more ways than one.
Given a German court just ruled that Google can be held responsible for the hallucinations of their AI search summary, this might not last long.
Re:
A great article on the ruling
Re: Re:
Yep. Read it; was good. Thanks!
Re: Re:
It’s a very good article.
One thing it doesn’t mention — and probably shouldn’t, because it’s outside the scope of the present case — is that this garbage product from Google is easily manipulated (a) because there are no humans in the loop (b) the crawlers behind it are vacuuming up anything and everything and (c) attackers (such as SEO scammers, phishers, etc.) have developed working, scalable methods for exploiting (b). (I trust everyone is aware that some of these people run custom web servers that serve up different content to different crawlers/networks/operations. You visit a web page, you see one thing; Google’s crawler visits the same page and gets something else entirely.)
Which means if, for example, I wanted to cause this mechanism to say that Heart of Dawn wears Crocs with glitter (thanks, Weird Al) I could probably do that rather easily and inexpensively. And until it surfaced, and until it was noticed, and until it was observed to be completely fabricated, nobody would detect it.
Google has built a machine not just for telling lies, but for telling any lie that anyone can trick it into telling.
Re: Re: Re:
But it’s something Google needs to be worried about. Imagine that you someone an AI to spread a lie about you, so they could then use this ruling to sue an AI company.
Could any of the AI companies even detect who was responsible ?
Could they prove it in court ?
Re:
You were right to call that out! Germany, business-killing anti-freedom socialist hellhole that it is, presently makes it much more difficult to recommend ingesting hydrochloric acid than do some other states.
I covered this back in May (https://www.freezenet.ca/google-is-shoving-more-unwanted-ai-into-your-face/ ), but this has long been a worrying threat to my site. I’ve already seen my traffic take a considerable hit when AI Overview was first introduced. Went from having 500 – 750 views a day to 75-100 clicks per day when that hit. In the months since, the site has gradually recovered and getting anywhere between 800-1200 clicks per day likely thanks to several projects and experiments. If this rolls out fully, I’m probably going to see another massive hit to my site’s traffic.
I’m hoping that this doesn’t actually roll out, but if it does, I’m terrified just how much traffic I’ll lose after all the hard work I went through to bring that traffic back.
Anything to get around the link taxes
Weakens their fair use argument
This seems to significantly weaken their fair use argument for the unauthorized copying of information used to train their language models. I have thought that the companies training language models using others’ data have had a pretty decent fair use argument so far as they are certainly engaged in transformative uses of the data similar to what was successfully argued in the Google books case. But now that the same company that engaged in the copyright violations is changing some of their other practices to more directly compete with and undermine many of the copyright holders it seems like Google is undermining their fair use argument with respect to their impact on the market, one of the other traditional tests of a fair use argument.
And this crap just reinforced why I stopped using google five plus years ago.
This is really well written. Just wanted to say thanks, I think this is the best article I’ve read on TD covering one of the downsides of AI, and it does it front on and fully fleshed out, rather than as a side note.
One other small downside I would add: because it’s used so much, there’s a huge cost pressure to use a cheap to serve model. This is why the existing Google search AI seems so bad compared to something like Claude- it’s effectively been lobotomized to keep costs down.
I also wonder how we go about starting to try to fix this, it seems like the incentives are aligned against it.
Maybe because I’m old or maybe because I’m autistic or both, but I loathe “natural language” interfaces.
With the old Google search you could type “Streisand effect” inurl:Techdirt.com -“beach house” and get the exact results you were looking for.
I very much do not want to have to type (or say) “Google, give me a list of articles about the Streisand effect on the techdirt website but don’t include the articles about Streisand’s beach house”. Jesus Christ, I got bored about halfway through typing that.
Re:
That kind of search is hard to do anyway, because of all the articles that aren’t really about the beach house but mention the history of the term in either the article or the comments.
Most people will not bother only because the answer is faster than clicking on the first link, skipping ads and newsletter popup, muting autoplay and scrolling to the start of the content.
What’s great will good webpages is that there is a lot of information about any subject, it’s not just few sentences summarizing what people could naively ask but the whole introduction, chapters and conclusion filled with many things people visiting the website wouldn’t asked but will still find very interesting.
AI chatbots are more the introduction in any Wikipedia article, answering when and why, but then there is also many chapters explaining how. For many questions, chatbots are neat since it give pretty relevance answers (sometimes wrong), but never tries to hint there is much more to learn about it.
As my grandpa used to say: Don’t ask “why” but “how”.
Re:
If I click 10 search results, probably 2 or 3 will be demanding that I prove I’m a human (as if I’ll remember that links 3 and 9, specifically, looked like they were worth putting up with some bullshit) if not blocking me entirely.
If one switches from “ai mode” to “all” you won’t have to take to the chatbot. The “All” tab still has the ai overview of course, but you’ll be able to scroll down and find the usual page of links.
Currently, the “All” tab remains the default. As long as it stays the default, there shouldn’t be too much of a difference.
It’s not so clear to me. Anyway, Wikipedia’s image-loading has been broken for months now—usually about 80% don’t load, and if I want to see them I have to click them and probably click the “Original file” link—and that’s not gonna encourage people to visit.
If I can get the text from elsewhere and the images are missing anyway—and it’s not practical for me to edit, because anonymous edits have been blocked for decades—what’s the point of visiting? But, also, how does my lack of visiting hurt Wikipedia?
Re:
Wikipedia relies on visitors to donate money to keep the site running.
Fewer visitors = less money coming in.
Re: Re:
They don’t, actually; or, at least, it’s far from clear. They spend about 3 million USD per year for server hosting, and have like 300 million in the bank.
They give a lot of those donations away as grants, and also use a lot to pay salaries. It’d be fair to say some salaries are to “keep the site running”, but probably most go beyond that (e.g. for “improvements” or just overhead). They’re at the point where an endowment, using assets they already own, would be roughly enough to maintain the status quo.
But also, since we’re talking about Google specifically: “The Wikimedia Foundation, which operates Wikipedia, struck a deal with Google in 2022 to have the tech giant pay for training access to Wikipedia content”. So, Wikipedia are already being compensated for the drop in visitors attributable to Google.
It’s not only search that has been seeing a steady degradation of quality. I use that gesture typing method on my Android (swype was the name at some point if memory serves) and part of it relies in predictions of the next word to make it faster. And boy it got so much worse in the last few months. It used to learn my gesture particularities and my pattern of speech thus providing better results and less corrections for gestures that were in some sort of intermediate between many possible words.
It’s clearly using AI now and I’d say it’s dipping its toes in some common training data because it simply stopped offering useful predictions AND it’s struggling to identify my gestures correctly with a much higher frequency than before. All that while refusing to learn words I use shitloads of times but aren’t really part of the lexicon like the names of my cats (before the keyboard learned them and could identify all, including one that had an indigenous name but now it simply doesn’t “memorize” anymore”) or silly, made up words we use as nicknames and such.
It’s happening across all products. I haven’t really noticed because I use GPS for reference and don’t really follow routes on a daily base but some people say Maps is tracing worse routes as well. Google is dead. It is now a giant zombie eating the internet we knew and leaving a wasteland where it touches.
Thankfully the old, rich internet is resisting in open initiatives. But overall these are sad times for tech enthusiasts.
Google AI doesn't know Transformers
Here’s my own example of how Google AI is now getting it wrong:
I asked Google which Transformer turned into a 747 aircraft. The AI answer was Jetfire, with a 4 paragraph explanation about it. Yet I know that Jetfire absolutely does not turn into a 747 AND that the photo next to the paragraphs was not a 747.
What happened to a direct link to Wikipedia or a Transformers Wikipedia, which both have the correct description and history of the character?
I reported the result as infactual, but I wonder how many people will think this is correct information?
Meanwhile, medical professionals are using “ambient AI” to take note like a court reporter during patient visits.
They suck balls.