Google's Results Are Infested, Open AI Is Using Their Playbook from the 2000s

Just sharing an anecdote:

I was searching for a quote that I'd heard in an audiobook the other day. I just had the general paraphrase, and didn't feel like scanning through the chapters to go find it. This was a somewhat obscure source.

Google had just straight garbage for me. The quote was political in nature, and I felt like the results were fighting general tone-policing filters and were tuned for recent events.

o1 on the other hand, found the author of the quote, summarized the general idea of what i might be searching for and then cited potential sources.

It's just patently obvious to me that google has failed in delivering the core value prop of their product, they're begging to be replaced.

SEO ruined the web, guided by Google's ranking algorithm.

Things will get even worse as scammy companies start flooding the web with LLM generated content pushing their products to bias LLMs to increase the probability of outputing their name for keywords related to their business.

Don't bet on AI staying clean.

A lot of HN readers conceptualize the forces attacking the integrity of the search results as just some isolated people taking occasional potshots, and then maybe slinking away if their trick gets blocked.

It is probably a lot more accurate to visualize the SEO industry as a Dark Google. Roughly as well resourced, with many smart people working on it full time, day in, day out, with information sharing and coordination. It isn't literally one company, but this conception is probably a lot closer then the one in the heads of most people reading this. Dark Google is motivated, resourced, and smart.

And then, once I started thinking of it that way for this post, I realized that increasingly.... Google is increasingly at beck and call of Dark Google. They're increasingly the real customers of Google and the real source of money. It's why Google just seems to be getting worse and worse for us... it's because we're not the real customers any more. Dark Google rules.

And if Dark Google has not yet figured out how to scam AI... it is only a matter of time. Dark Google is where Google gets its money now. When Dark Google turns its attention to AI fully, OpenAI will be no more able to resist its economic incentives than Google did.

Can't wait for the first screenshot of someone searching for the impact of the battle of Gettysburg on the civil war and seeing the AI do its subtle best to slide an add for Coca Cola into it in some semantically bizarre manner.

> Does ChatGPT Search have trust? Open AI isn't monetizing its search just yet, but AI has its own issues with hallucinations.

Everywhere where SEO people congregate, they talk only about this: how to produce content that will eventually end up in training data for LLMs, so that when you ask about anything remotely connected to a given brand, its products will show up in the response.

Ads are bad enough today, but it's possible the future will be worse: product placement in everything, everywhere, all of the time.

A lot of folks are focusing on the AI answers that Google gives, but for me the real downgrade has been the change in their algorithm a few years ago where it tries to search for what it thinks I mean instead of what I search for.

Even putting entire chunks of text in quotes isn’t enough anymore. I can never get Google to search for what I want without trying to engineer a prompt, when it could at some point.

In trying to become more helpful, it’s become worse.

The Yahooification of Googles search becomes more amusing when one considers that it is now led buy the dude who ran Yahoo search to the ground.

> Enter 2024 with AI. The top 20% of search results are a wall of text from AI...

I'll be the contrarian here and say I actually like Google's AI Overview? For the first time in a long time, I can search for an answer to a question and, instead of getting annoying ads and SEO-optimized uselessness, I actually get an answer.

Google is finally useful again. That said, once Google screws with this and starts making search challenging again, as it has been for years, I'll go elsewhere.

Google in 2000s was excellent. Modern Google makes me feel like I’m a product being sold a lemon at a barb wire fenced used car lot. It’s horrible, the things being shown are horrible, and there’s questionable ethics and value to be had by even going there.

I remember when the more tech savvy folks were migrating from Altavista and Yahoo to Google because they understood it is better sooner than others. The same thing is happening. I consider myself tech savvy (still!) and I have to admit that I rarely use Google for all sort of information.

As a side note: I am using Safari and I noticed that Apple's search is also replacing my Google searches. In the past if I knew name of a company or organization but not their website I'd Google it. Now I put it in the address bar and Safari very often finds the website for me.

The question of course is how well that level of quality will hold up now that more and more of the internet is AI generated, and said AI generated content is being sourced for said AI tools. It feels like our choices are either to only get information from a certain time period or earlier, or to accept the information provided by Open AI and co is only going to get worse and worse over time...

In 2000s we were freeloading on the investors expense, in 2020's investors are recouping their investments.

Also, most of the content was either stolen(divx, mp3 etc) or created without of expectation of immediate reward(mostly passion projects).

Oh and btw, Google didn't got infested after LLMs proliferation. Google results were useless way before that. With LLMs there's even improvement as the spam is at least mediocre content.

There is no cure to the endless stream of AI generated SEO-trash. As long as there is a system to game and an incentive to do so, it will be gamed.

The only solution I can dream of is to remove the incentive, aka remove advertising. I'm afraid I'll be dead long before that.

OpenAI trains on sources like BusinessInsider which allow sponsored content for a price. This is how they will monetize their products, by intentionally introducing it to bias for a profit motive. It's a good thing they convinced you that its a black box, otherwise you would see this coming.

I'll be honest, I don't understand the example in this article -- it doesn't seem to be evidence of the thesis/complaint. The two dark mode screenshots are basically the same and, at a glance, I trust the Google one more because it's showing me a snippet of the authoritative source, Apple.com. Beyond that, there's no ads in the Google screenshot, which was a huge part of the thesis statement.

> OpenAI's search is becoming Google in the 2000s

Google started going bad in the 2000s (albeit not as bad as now).

> if it can remain trustworthy.

At no point was it trustworthy - even if it were an abstract LLM, trust would be an issue; but this is the opaque product of a corporation heavily invested in by untrustworthy entities and people.

That does not mean it isn't often useful, but "trust" and "usefulness" are two very different things.

In 1999 I worked at Yahoo! It was great the stock was flying, I worked on really cool tech writing C++ code. Then one day I went to a social dinner and a high schooler got up and made a speech and at one point said “I googled it”. Right then and there I know the gig was up.

So far I have not heard anyone say I GPT’d it, but Google is running very dangerously close to the edge here. For one thing the founders have checked out, never a good sign.

Something that also bugs people is GOOG wants to follow you everywhere, when you sign in to many websites that little blurb asking for your google account comes from a google server (<script src="https://accounts.google.com/gsi/client" async defer>).

I was responsible for servers that ran 100m page views a day at Yahoo! One day I was approached by this smarmy little guy who asked if he could pull logs from the machines. Alarm bells. Who the heck was this and what was he doing with the logs. I knew of course he worked for Filo and so I had to give over the data. This was the start of the spying on the customers. Google is a master of this, and it really irks a lot of their customers. Another red flag.

Alternates like duck duck go and brave have made some inroads. Their percentages are quite low still.

There have been layoffs in the name of cost cutting. Googlers have had some very public employee dissatisfaction meetings (my name for them). Employee compensation problems, problems with businesses the company is etc.

One last thing, Mark Cuban sold Broadcast.com to Yahoo for $5.7 billion April 1, 1999 (seriously!). “ Apollo Global Management acquired a 90% stake in Verizon Media, which included Yahoo and AOL, for $5 billion. Verizon retained a 10% stake in the new company, which was rebranded as Yahoo upon the deal's completion.” The deal was finalized Sept 1, 2021 according to chatGPT.

Definitely agree, but I'm surprised Perplexity wasn't mentioned in this post. It's currently Perplexity Vs. Google

This! "I don't want to watch a 10-minute video for a quick answer."

And this: "OpenAI's search is becoming Google in the 2000s, if it can remain trustworthy."

The problem I see: People use OpenAI/Perplexity for knowledge. Not to seek website. I think sooner or later, most website will block AI crawlers. What does a website gets out of it?

So... the next step clearly is ChatGPT adding Ads to its output.

Of course OpenAI can't stuff their UI with ads yet. In the middle of escalating anti-AI sentiment due to the rapid slopification of the infosphere that they intiated? With cultural resentment growing over the devaluation of visual evidence? With video and audio modalities creating distrust and obsolescence in the creative class?

Google only had to provide superior value with a clean UI. OpenAI has to contend with normalizing the mechanisms that are upsetting the lives of the customer base that pays them; the customers they'll replace with ad servers as soon as it becomes prudent to start indicating their end-game.

So, as a company, when can I pay OpenAI to twist its responses towards my company?

Because it will happen. I’m afraid that ChatGPT is my friend for 20€ today, but prices will increase and response quality will go the way of Siri.

Today I had a better experience with Google than with ChatGPT.

I started with chat and asked why I couldn’t change the passcode on a kid's device.

I tried 3 answers. None worked.

So I google. First hit is an Apple forum with the exact problem. Solved.

I think I’ve had such a bad experience with openAIs charGPT answering a question falsely and sounding very convincing that I skip the AI overview or double check to confirm it’s correct. Am I mad?

Trying to replace Google w/GPT but....

- Local search results for businesses, phone #s, directions isnt available

- Im paying GPT subscriber and it constantly logs me out on my different devices

- Sora I can not upload photos of whoever & make them do whatever like https://hailuoai.video/ which is free and at times fairly convincing (fun for the 12 yr old in you lol).

I know GPT Search is new and Im excited for GPT to become a phone AI OS or Open AI & MIcrosoft developed their own phone with a new personal device paradigm... i.e. create a H.E.R. phone.. it's your personal AI Assistant that does all for you via text, voice, hand gestures, facial expressions, etc. Once you pick up your phone you see your assistant waiting to assist. You can skin your AI assistant to look like anyone living or dead (loved one could live on & help your throughout ur day). Probably some crazy ideas but a H.E.R. phone / personal device as described (some parts) would be something new/different and possibly give Apple and Android a run for their money!

People want to spend less time on a task. Whichever tool succeeds at this will win.

I am working with a new for me tech, and got stuck integrating an SDK from a hardware manufacturer into my runtime. The SDK is in C, and my runtime isn’t.

A friend suggested to use his LLM wrangling skills to get the SDK adapted. The results were interesting, but we wasted a day or two on trying to make it actually work and failed.

Then it turned out I can call C code from within my runtime, which the LLM did not point out. I figured out the specifics with a bit of googling and successfully managed to call C functions. Yet there were still issues in parsing the data, now in C SDK itself; the friend used different LLMs to refactor and comment C code, but they did a bad job with missing implementations and a lot of general obvious comments (an empty deinitialization function with “clean up as needed” comment and so on).

Then it turned out there was an SDK for my runtime. I just did not search enough. Of course, the LLM would not say that. The LLM would just obediently try to do what it’s asked to do and never question why.

> And even these ads weren't that bad. What made Google successful was showing ads you wanted to see. I'm searching for a bottle of wine, and ads for bottles of wine were shown to me.

Even at that time, Google hardly ever showed me ads that I wanted to see (even though in former days, Google's ads were less intrusive than they are today): as I sometimes mention on Hacker News, most advertising networks have difficulties to "pigeon-hole" me into interest groups - many interests of me are somewhat niche.

For example many Google searches that I do are about scientific topics from my areas of expertise - this is a rather hard topic for advertisisers (though it is plausible that it might be a lucrative topic for anybody who is capable to monetize on people like me).

On the other hand, if (rather accidentally ;-) ) Google's advertising/interesting sorting algorithm actually "finds" some interest of me which can be "monetized" via ads that try to sell me something, it's the common case that I have already bought such a product recently - in such a situation I clearly don't need another piece of the product. The reason for this is simple: after I bought the product, I have to understand some of the product's "complicated convoluted details", so a lot of Google search queries are of the type "[product name] [potential problem that I have with the product]". Thus, I do a lot more Google search queries for the product after I bought it.

It should be obvious that for the ad to pay off for the advertisier, it should be shown before I buy the product - thus Google has to be capable of interpreting the quite weak signals that might show that in the future I might want to buy such a product, and not the strong signals that Google sees in the weeks after my purchase.

This is correct. A ton of the hype about LLMs is you can type in a question and get a direct answer without needing to struggle through listicles. Phenomenal compared to the trash a search result provides.

Is is ironic that LLMs are the source of much of the garbage in search results. Good business model to produce unbearable noise and the filter that recovers some signal, I guess.

Google seriously needs to scale up their generative models to all of crawling/indexing/ranking infrastructure. Their current ranking models are not capable of dealing with the next-gen web filled with 99% gen AI craps. I think they also know this. The problem is the cost and they're hyper-focused on bringing it down, but it is not fast enough.

I don't think it will be as easy to supplant google search as it was in the past with Yahoo and Alta Vista. In the past internet search was just starting we were all used to trying different search engine and the results were different enough with all the different ways of indexing and searching that moving from 1 to another made a difference. Most of the people using the internet were a bit of pioneers navigating through something brand new and unique.

Today majority of the internet users ie 10-32 years old or 60%+ of the internet users grew up with using Google and how to get best answers our of it. Chat gpt might bring in some churn but as long google is close enough it won't get replaced easily.

> When Google came onto the scene, I credit its success to the tried and true paradigm that makes companies successful: simple and easy to use.

If Google's competitive advantage was only that they had a clean and simple web page, I don't think they'd have nearly the sticking power they ended up having. You can design a front-end that looks like Google's original UI in a matter of hours.

It seems implausible tale that in the 25 years since their ascendance, nobody has tried to compete with them with a simple-looking search box.

Google became dominant because Google was much better than the alternatives.

The example he shows here has the answer to his question highlighted in the AI overview, and in huge text in the InfoBox. He got his answer instantly without having to go to another page.

Really wish such bold claims had better evidence.

Google's results from the AIStudio are really good results. Everything there is good. Probably because the studio has better models than are currently being used for general search.

AI is just doing what SEO people have been doing for decades, inflating results that have no business being there. At least now there's no pretention of having any skills.

Once timeless facts will be solidified in AI training, like historical dates, laws of physics, maths formulas or specific programming API, there will be no reason to search for these things using Google.

The web battle will then happen on this moving quickly, like the news.

This will give immensely more power to the medias, and I fear that a lot given they have demonstrated time and again they can't be trusted with it.

In addition to the ChatGPT’s search, the advanced voice model also supports web search, making it even easier to search.

I wouldn't be concerned about trusting the results of ChatGPT if it also were providing links to the sources it had cited or used as a reference in its answers.

Unfortunately, it doesn't, and so I can't verify them. Not sure if it's an actual limitation of current LLM or rather they're intentionally filtering out the sources.

I was using ChatGPT to compare Docker and Podman and getting reasonable comparisons. I also asked it about c code searching tools and getting a reasonable list with what I think were reasonable comparisons.

It hit me that in a few years, this may not be available as Docker and other tool suppliers start paying for advertising. We’ll see.

I may be overly cynical, may live in the rarified bubble of the tech-savvy, but I think Google as it has existed is already dead and doesn't yet know it.

Their search results have been less than useless to me for some years now. And then LLM's came along and have been my go-to for all queries.

Investors value OpenAI more than $150bn, but it just reported $5bn loss on $3.7bn revenue: the financial pressure is tremendous and that cannot last long. It's the same kind of situation as with Google vs. Yahoo in the early days, when the clean interface without much ads was a very strong USP.

This. I think Amazon is doing the same thing and leaving the door open for Walmart or another player with some logistical muscle to take a bunch of the market from them. Amazon search used to be so simple. Now you sift through so many ads and hidden ad garbage to find the thing you want to buy.

Will OpenAI eventually deliver ads along with responses?

“Before getting a response, a word from our sponsors:” type thing?

> I don't want to watch a 10-minute video for a quick answer

Nailed it. Google should show the transcript.

All I see on Google for most searches is ads and SEO spam that gets the ranking and clicks. It’s only once I spend time scrolling it that I realize it has the telltale signs like very lengthy text that doesn’t answer the title succinctly.

Infestation also applies to actual "organic" results. The old Google philosophy would have ranked pages with annoying pop-ups, paywalls, sign-in walls etc out of existence. Today's Google does not. They even tolerate sites that show them articles to index but are then hidden from the general public behind a paywall.

The old Internet still exists, but Google's ranking behaviour hides it.

> Google is losing trust with all these buzzing results, and its answer is to throw more jam at the wall to see what sticks. But this just attracts more flies.

If you're going to work the metaphor, work the metaphor!

> When Google came onto the scene, I credit its success to the tried and true paradigm that makes companies successful: simple and easy to use.

> Yahoo was dominant back then, and it tried to put everyone and everything in front of you. Then we learned about the paralysis of choice. Too many choices, the mental fatigue weighed in, and the product became difficult to use.

This nonsense again? I was around then, and I switched from Yahoo and AltaVista to Google despite its dumb name and stupid, childish logo because Google's results were hands-down better. Instead of a solely full-text search paradigm based only on keyword density, Google also ranked pages based on how many other pages linked to them, the so-called "PageRank" algorithm.

This worked much, much better, and was much harder (for a while) to game. Before Google, it was common when searching to find pages that gamed the search engines by stuffing their <meta> keyword tags with SEO crap or putting it in giant footer sections in a tiny font the same color as the background (to render it invisible). Google's PageRank wasn't fooled by this.

Also most of the major search engines adopted similarly minimalist UIs, and it did zero to stop the bleeding. They all lost to google. (AltaVista, the pre-Google Google, was still useful for a while for some specialty searching, like for anonymous FTP servers, and I wonder if DEC had never gone under or if Compaq had spun off AltaVista, maybe history would be different.)

EDIT: I just realized the article doesn't even mention AltaVista. Unbelievable.

I actually think this might push me towards using ChatGPT more. I just hate having to type, "What version was the iPhone in 2016?" rather than "iphone 2016 version number" lol

Google became an asset of the government. By selecting what you see they control what you think, what you buy who you vote for, similar to legacy media.

Open AI has started on day one with that goal.

Between Kagi for ordinary web search and Claude for “complicated questions I don’t know how to phrase for a search engine”, I’m pretty happy right now.

Great job to both organizations!

I have been using Perplexity 95% of the time this past year.

Also check out Grok, it's surprisingly good, especially at organizing the results to get the job done for you. You can access it as part of X.

I'm happy to pay for ChatGPT, and I would happily pay for a clean Google search experience without ads. How much are they making per visitor a month?

Brave Search might be a good alternative.

It's index is created by people surfing with the Brave browser, so only websites used by real people are included.

I'm surprised nobody (in the comments I scrolled by) has brought up the Web option (which only shows links) yet

Google has put it's early 2000s history where it can't find it! In the garbage most likely, but somewhere deeper.

This article woefully assumes OpenAI has had the same R&D that Google has had as a search business.

chatGPT offers bullshit answers faster, and more confidently. However the issue for openAI is the cost of business is _horrifically_ expensive.

Sure they charge some users for premium access, but they aren't currently enough to cover costs.

openAI needs a step change in performance, and that meta doesn't release an open-source version of it and a step change in compute efficiency.

I honestly don't mind the current Google Search Results page design and additional features.

For me, it's the web in general that isn't the same. Google Search is probably just as good, but I have an idea that the content to be indexed isn't what it was.

Has the world changed, or have I changed?

Probably both.

The WWW of content isn't the same, and i don't search for the same things anymore.

My use of search has become integrated in my patterns in the physical world. I need to know what products are in stock, where things are, when a restaurant is open and so on.

For those searches, Google is generally excellent I find.

Also, I actually find Google's AI Answers to be pretty decent.

Even if I do ethically disagree with denying the original content authors those page views.

What is the current status of SEO weasels trying to poison AI data by stuffing it with spam?

Agree on the idea, however I think Perplexity is a better product than Open AI in search

google's playbook from the 2000s was to run a search engine based on a novel ranking algorithm that actually worked with a ui designed by engineers who like things fast, simple and minimal.

the rest of the process described herein has been best described by cory doctorow as "enshitification."

i’ll buy this. seems reasonable.

Why OpenAI in particular?

I mean, Gemini is much better than the "traditional" google search as well.

Anything is better than 2024 google search actually.

I for one love the “wall of ai” this author decries. It is super helpful for most of the questions I ask. I don’t know why he doesn’t like it.

yeah Google Search is awful and there's still nothing better. What's new?

Still using Google I see? How primitive..