I think this is a step beyond that - why should people be creating cooking websites when you can ask LLM how to cook given thing, while indeed, serving their own ads. It's the continuation of "we own content other people produce" policy

there are other such goodies like mashed potatoes with broken lightbulb gravy, or fiberglass omelette, enjoyed by beldar conehead.

i wouldnt trust an AI for any recipe that i dont have personal experience with.

the safety rails are not very strong yet.

Point is, unless you're stupid enough to add glue or broken glass to your meal just because a recipe told you to, it's perfectly safe. More than just safe, LLM recipes these days are utterly boring in their normalacy, and, unlike cookbook recipes, can dynamically adapt to what you actually have in your pantry.

I could see LLMs being helpful to explore what's out there, like finding similar dishes or dishes involving a specific set of ingredients or dishes involving a particular technique, but a pretty poor tool for the actual technicalities of cooking or more importantly the uniquely personal aspects of food culture.

I dunno. I'd just buy larousse and on food and cooking.

Food lab, for example, covers buying storing and cooking lamb + a guide for a 5-7lb boneless leg across 5 or so pages. Kenji goes through great lengths to build intuition. I'm sure larousse, which is more of an encyclopedia, covers lamb quite extensively but it's probably more terse.

The internet can be an excellent source, but like most things it depends on who is writing it down.

I'll stick with my mom's handwritten recipe book.

The computer generated recipes can be useful as inspiration, but of course common sense is required.

"On two occasions, I have been asked [by members of Parliament], 'Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?' I am not able to rightly apprehend the kind of confusion of ideas that could provoke such a question."

~ Charles Babbage

Okay, so you don't trust LLM, so you go to a website instead. And... LLM-generated pages are SEO'd to get the top links. So you can't trust any website now (shoot, so much nonsense even before LLM, just more obvious to some of us). So basically everything on a computer is untrustworthy, directly from an LLM or not, unless you got yourself a copy of Encarta '97.

So you pick up a book at the local library. Librarians picked some books to order in subject matters they aren't expert in. How do you know those are accurate and safe? If the book says to use rice starch glue, how do you know the author didn't just copy that from an LLM? Or make it up?

Trust is fading entirely.

The thing is, these things do produce some kind of result that looks like what you want. But it is still up to you to test these things on a project before you rely on them for whatever it is you really wanted them for, and that requirement doesn't go away just because you sourced the information from some LLM, or a book at the library, or Nick Offerman, or whoever else.

AI got better over the last couple of years, and you didn't keep up, and because that's not going to stop, it will eventually become a problem for you.

> because that's not going to stop, it will eventually become a problem for you.

How? Will it stop being possible to cook without AI?

That's like saying the fundamental technology behind an Egger-Lohner Hybrid and a Prius are the same. Technically true, but if you use that truth as a basis for decisionmaking, you're doomed. A modern AI model wouldn't make such a foolish mistake, so you'd better not make it yourself.

https://cdn.bsky.app/img/feed_fullsize/plain/did:plc:allu5vs...

Aside from that, please let me know when you find a machine or a human that never makes mistakes. I'd like to invest.

There isn't a robot straight up putting glue on my pizza. I will be following the recipe myself and since i trust myself to accurately detect that glue shouldn't belong on food i have zero issues with doing this.

Can you really though? Are the results delicious? I've never tried that.

That being said, an LLM can give creative ideas, mix and match components, but you should not trust the details at all.

I wouldn't just ask it to make me a "novel and interesting recipe" giving it a bunch of weird ingredients that don't fit together. It would probably try its best but garbage in garbage out also work for food! At that point i ask it what should i get from the shop to make a recipe with x ingredient that i have in the fridge and i want to finish.

It's also pretty good with meal planning if you do that, it can estimate portions with calories breakdowns etc...

You can - but it's not advisable, not in the least.

And I think the rest of it will end the same way. People will be significantly less eager to do all that free work when no one will be able to find it.

They believe they won’t be wrong for long.

Block Googlebot from your sites.

Let's go back to webrings.

They killed their competition and now they will give you the product that gives them the most money.

Though that's not to say they're acting altruisticly here.

Google seems to be racing toward a new dark pattern where users learn to trust rely on the AI for neutral, smart objectively correct answers — which boosts trust in its sponsored product recommendations. Super gross.

Not me. I really appreciate having both results simultaneously. I can scan the first couple sentences of the AI response, and if that already has the answer then great. I can expand it to see if there's more.

Or, if I see that the AI mode didn't understand my brief search query, I just glance at the search results below.

And often times, when I do need to follow a link, I find the source result links in the AI mode to be a better quality than the search result links.

It's the best of both worlds.

But how do make the determination that the answer is good and you should stop reading the page? Vibes?

Most of the time I'm looking for something very specific that there are plenty of articles about, but clicking on the articles results in popups, banners and an unhealthy amount of scrolling to get to the answer.

AI overview provides me the answer instantly.

Think about suff like "does china borders afghanistan". In those cases you can be confident that the AI overview is right, and saved you time.

If it is a complex or niche question I tend not to trust the overview and go straight for legitimate-looking results

Well before Google screwed it all up there used to be some correlation between top hits and what you were looking for. SEO has muddied the waters for many years now and it’s never been truly “merit based” or “objective” or whatever we want to call it, but generally speaking, the first results were the best by default.

SEO optimization totally ruined google search for me for the past few years

Ok, but it’s been mine. And clearly I’m not alone.

I feel like at this point any discussion about LLM’s has an implied “my experience” because LLM’s are super inconsistent due to not being refined tools at all. I’m sure your experience has been different, just like my experience has been different. I imagine you’ll want to chalk it up to operator error, but it sure seems like a lot of people have variations of my experience. If so many people are operating it wrong, then maybe the tool is poorly designed.

Understand that I use LLM’s pretty frequently. I am not “anti-AI.” I’ve used production tools incorporating machine learning for years now. But LLM’s simply aren’t the bespoke tools that these companies want you to believe, and they are definitely not a suitable replacement for search. It’s simply too inconsistent and will hallucinate answers. Google search didn't make up answers, it presented indexed sources that you ver in real time which I find to be a far superior way to do research. I don’t like having to guess when an LLM is just making shit up as it asserts something with simulated extreme confidence. Not only that, you can take a correct answer from a LLM and just start saying “know that is not right,” and it will start apologizing to you and generating other answers. That is a huge problem! I shouldn’t be able to “convince it” to give me a different answer.

Yes SEO made things objectively worse. Doesn’t mean we need to add another layer of issues on top of that.

The same way I make the determination as to whether a linked search result is good and I don't need to click on another search result.

It's not like non-AI webpages are inherently more trustworthy or anything. The internet is full of misinformation everywhere, you know?

Users aren't adopting their AI at the rate shareholders expect, so they now force the adoption at the cost of search.

>Just one year after its debut, AI Mode has surpassed one billion monthly users, with queries more than doubling every quarter since launch. As people have realized just how much more Search can do for them, they’re searching more than ever before — so much so that last quarter, we saw queries reach an all-time high.

>Another place where we’ve been rapidly innovating is in the Gemini app. Last year at I/O, the Gemini app had 400 million monthly active users. Today, we’ve surpassed 900 million, more than doubling in a year. In that same time, daily requests have grown over seven times.

If the results are garbage, or people have difficulty with it... Of course number of searches goes up. That doesn't mean the product is better or its not resulting in brand damage.

Of course they have more AI queries every day. They have full control over what goes to LLMs and what doesn't.

I'm not entirely sure I'm buying that, unless users keep prompting the AI to reduce the amount of reading they need to do. Sort of interrogating the AI, rather than reading a Wikipedia page.

People have many more questions in their life than they do queries.

Most non technical people I know asked questions to Google even before the AI overview. Instead of looking for the answer in seo-bloated articles, they find it in the overview.

I think google should improve in detecting the kind of query when I need a link that I don't remember, and deactivate the overview on those. If I search for "ryanair booking" I clearly need the url for booking a Ryanair flight, AI overview is useless

If programmers and engineers are saying "why would anyone want that?", odds are the product will be a gigantic success.

Rather than "users don't find what they want with the AI as easily so they have to spend longer with it"

No, it's not. AI mode is something you have to select (in the search window). There is an AI overview provided with your basic search results.

That's what I get, in the UK, logged out of Google, from a search in Firefox omnibar using "Google" as provider.

I'm aware that they have other things that can be described as AI modes.

AI Mode in that screenshot is the tab to the left of All.

I wonder how much the search results thing is related to language and locality. I have a hunch but I haven't really dug into it.

I live in the US, I speak English, and my browser is normally chrome.

The number of times I've gone to the 2nd page in Google search results you can probably count on one hand in the last 15yr or so.

I use the standard Google search things when I want specifics... Using quotes, site:news.ycombinator.com to search a site, or add a "-" to remove results from that site. I use a "+" when needed. Nothing fancy.

When people say they can't find things in Google search, I'm genuinely baffled. I have a strong suspicion that it has something to do with the combination of browser, locality, and language. Why? Could be tons of reasons for that, some probably anti-competitive on the browser side.

I have tried to use ecosia, start page, duckduckgo, etc. Was never happy with those results and always ended up back at Google search.

I just want to know what's different, you know? I look up some pretty obscure stuff sometimes.

Note: I do normally have my Google account logged in in the browser when doing search, however I have search personalization and history turned off, so that should not be influencing the quality of my search results compared to whatever "baseline" is.

Regular people are/were really bad at using google, so google moved towards showing what it thinks you want rather than what you want. They paved over the skill gap between people who understood keywords and word order, and people who just typed in a quasi legible sentence to find something. In doing so though, they killed a lot of skill that people had developed with google for years.

Basically they made the game worse for pros so it could be better for amateurs. I have never heard a non-tech person complain about google getting worse over the years, and they seem to overwhelmingly use AI overviews now too.

Hard agree. The only thing I've ever witnessed another person do on Google (this is only an incredibly slight exaggeration) is:

1. Type a 'query' - either a brand/website name or some kind of stream of thought like "dishwasher error 03F" (without quotes)

2. Click or look at the very top thing in the results.

This used to mean 80% of the time they'd click the top ad, 20% the top organic result. Then they started putting non-clickable "answers" in that top spot, which would always be accepted as 'the right answer'. When those appeared, approximately no one would ever click any 'blue links.' These started out pretty reliable because they were just direct extracts from sites like IMDB: "Brad Pitt is 44 years old" etc.

Now it's like 60% of the time an ad, 40% of the time their bargain-basement-model "AI Overview" slop. Either way, approximately all users always just use whatever is on top and ignore everything else.

Wtf

Citation needed. A hard push to change their search offering, sure. To improve it? Well, if by improve you mean 'require more interaction and viewing of more adverts on average before leaving' ...

I would bet all of my money that you never once did a Google search (pre-LLM mania, but maybe even after) that looked like

"What kind of clothing is best for when you are going hiking around the lake, so my feet don't get so cold?"

Sadly, this is how most humans have used a search engine for decades now.

And you frequently ended up finding a discussion forum with around that question and relevant discussion under it.

Forum hiking warm socks backpacking trail running

Which was basically a structured programmatic query that activated the old google algo just right.

I often use terse searches too, mostly when I forget to write it out longhand to satisfy Google - but either way it's getting less wheat with my vat of chaff in the SERPs and several times recently I've had to re-phrase to get anything useful.

Was I the frog in the pot and now I'm cooked? I don't feel like in search Google any different from maybe 2005 or so.

Technically, all the people who google "how do I disable this shitty AI mode in google" would count as "driving an increase in total search queries."

An easy way to make a feature popular is to force it on everyone. Then you can pat yourself on the back when 100% of your users are using it!

I search more when I cant find the thing I am looking for. I search less when I find the thing I am looking for.

Second, it takes additional effort to not do AI search.

Because the goal is not to provide the best answers.

It's for users to train their AI.

1. To protect ad revenue they make search results worse to increase the number of searches by making people refine their searches. This made people upset because search result quality went down. 2. "AI everywhere!" put them in a panic, so they shoved am LLM into results, hoping it could pick through bad results and give good data to the user. 3. LLMs are expensive to run, so they're using a cheap model.

Cheap model + bad results = abysmal user experience.

There are too many groups with opposed interests fighting. Ad groups wants worse results so people search more (not realizing this just drives users away). Search groups want a better product so they stop losing users, and the AI group is being given a bad name because management is using their worst AI product on search. So the whole experience is just garbage.

Why would this work? Were yahoo and askjeeves sandbagging their results too just so they can get more clicks?

Humans are predictable and hate change. For a short while it DOES work, people are used to great results, assume they're not using the the best keywords, and they'll reformulate their searches. For a while. After a while of all searches being not as good as they used to be, people start looking for other alternatives, which is why DDG is seeing an uptick.

It's called enshittification. It's easier than improving a product.

> Were yahoo and askjeeves sandbagging their results too just so they can get more clicks?

No idea.

The platform has been various kinds of hostile for a few years now

Slight digression: Claude/ChatGPT/etc all can search the web, but Google's AI already has a local copy of the web. It's much faster because of Google's TPUs, but also because Google has a copy of almost the entire web available locally. I recall others testing this and they observe that Google doesn't actually make HTTP requests to sites it references. It just uses its local cache. That's an advantage that all others seem to lack.

Of course, I agree that when I want search, I want search. But personally I've found if I want an LLM to very quickly answer a simple question, the type of thing all of them would do an equally good job on, I prefer Google's for its sheer speed.

"AI" gets higher volume of use than search. This was disclosed by Google under oath

More traffic, more usage time, more data collection

https://www.wired.com/story/even-if-you-hate-ai-you-will-use...

Like the parent I use good/paid AI when I want an AI response. So, yeah, an omnibox that knows when I want "an answer" and one that knows when I want to find a thing sounds slightly more convenient than switching between two tools, but Google search is not that Omnibox.

So now you're trusting them to provide the cure?

...and it really hasn't been for a good number of years now. I left a while ago when results were all SEO copy pasta blogs this is just a final nail in the coffin.

Because Google wants to kill off its search engine here. It is very clear.

> I assume it's a play to test the waters for how the ad market is going to work, because as a product I really can't see why I would ever use it. Dropbox comment moment incoming?

This assumes that Google search is still a high priority for Google. With their privatized adNetwork, they are trying to get people to trust them, and abuse users via their ads. That is their business model. Google is an adCompany. It stopped being a tech company many years ago already.

Also they control the adMarket for the most part. Just look at youtube.

Like you, I use both search and AI separately. Even casual, nontechnical users are starting to work like that. Including AI with traditional search results will keep a lot of users from jumping ship in the first place and will help win back users from ChatGPT.

I know a lot of people hate AI - at a minimum, there’s a vocal minority - but the reality is AI is eating search like nothing we’ve ever seen.

To each their own.

google pushed it into their other products to attract people to AI

there was and still are a decent number of people who haven't really used it, as crazy as that sounds

If I truly want to search I will ignore the llm results, but I like the convenience of a quick llm search that knows "all the things". I get the answer to my question without searching multiple ad-ridden websites (since the ad provider does all the things)