Expanding AI Overviews and Introducing AI Mode

Ads for products and locations still show up (at least in the existing Overview).

Ads in the AI overview. That's always been the endgame for them here.

There's no such thing as "the nature of AI Search", our current understanding and hope for what it should be is being projected, but as long as there's attention, ads can be there.

It's necessary to defend search, even if this cannibalizes search it's better they do it than let someone else do it.

What 8K TVs are the best ?

Definitely go with Acme, they're our highest paying^W^W best performer in this segment.

> because by its very nature it bypasses most of the ads.

for now.

> I really don't understand what their endgame is here.

To not lose. History is full of stories of incumbents not wanting to cannibalize themselves and dying because of it.

creative destruction.

They will kill their own product with constant innovation before someone else do.

The endgame is to increasing the organic look of the results. AI Search makes that possibly even though it is not reliable. Also, Google is probably under immerse pressure to "add AI" into to its products.

AI Everywhere is the new Everyone Can Code.

You noticed how no one talks about code bootcamps? LLM bootcamps will soon be popping all over the interwebz.

I appreciate the alternative perspective.

I have found them outright wrong or misleading often enough to generally just skip them, or get mad when I slip and catch myself reading them.

> I've seen lots of non-tech friends and family using them.

Same. It's because it's the first thing that pops up and they don't know any better, and they trust Google.

I've seen misleading or wrong information often enough to not waste my time reading them.

The downvotes confirm how HN feels but I largely agree with you. I do not find them wrong as often as everyone here claims and it appears most regular users are enjoying the feature at this point.

For me, it's the fact that content generated by an LLM is fundamentally different than content that comes directly from a search index, but displaying them alongside each other conflates the two. Most people don't know the difference, and place the same level of importance (or maybe even more importance) on AI-generated content. Yes, this content is convenient. However, if the content isn't accurate or correct (which it may or may not be, given that it's just a statistically likely sequence of tokens) then is it actually beneficial as a whole?

The failure cases for them are really bad. I've lost count of the number of times I've seen people "prove" something on social media with a screenshot of a Google AI Overview that's parroting false information it found on the web.

(Try searching for "encanto 2").

how is it easy to ignore? it occupies the space previously used by the top 3 search results

I think it would get a lot less hate if Google just made it possible to opt-out of them without having to download some third-party browser extension (which is actually what Google's AI summary itself suggests I do when I ask it how to turn itself off!). A small "opt out" link on the top of the AI suggestion would make most of this hate go away, and would make me stop paying for Kagi, if Google cared.

I also find it extra frustrating when AI summaries appear when I search for correctness-critical information, e.g., "what temperature to cook chicken?" or "can I eat old eggs?" --- why force me to scroll past an entire page of AI generated, 1%-chance-of-being-a-literally-lethal-hallucination "summaries" in order to find the CDC's actual recommendation? I don't want to play Russian roulette with my health hoping I don't get a hallucination, instead I just want the authoritative answer. Which Google did an amazing job at until a year ago, and Kagi is doing a pretty great job at now.

If I may suggest something crazy: Try Bing for a month. You can turn off most of the crap.

Backstory: I've been a Google user since it was in beta, having moved from AltaVista. A few months ago, I decided I wanted to try something different in an ongoing effort to reduce my Google footprint. Kagi is the obvious alternative, but they don't have the right plan for me — if their no-AI plan offered unlimited searches, I'd be using it. As an experiment, I set Bing as my default search engine, and it's been fine. You can turn off the front page "tabloid" crap, and also Copilot responses on result pages.

> Can I get back 2000s Google results?

https://kagi.com/

I can't say for sure, since I don't work on search or web crawling, but it doesn't seem unlikely that the Google of 2000s would serve nothing but spam on the internet of today.

> I find myself just asking Grok for things I typically ask Google.

Yes, and that's why Google is embracing AI here. "Web Search" doesn't perform as well as it used to, and people are usually looking for answers, not search results.

What I do miss, and I think what you're getting at here, is that there is a certain amount of dark matter in the web today – all the little sites that used to show up on google to fill little niches throughout the internet now feel undiscoverable, even if they are still there. Every results page feels like a list of forbes.com and medium.com news/advertorial/aggregator sites, a couple reddit links, and a couple of forums.

I can't, for example, ask Google for a list of online portfolios of artists who have experience in three.js. It seems like back then, you _could_ do that and get a reasonable list of results.

You can't. When Google started they relied on keywords and ratings based on # of linkbacks. Then SEO optimization became a thing and people started to add (fake) keywords, and generate (fake) linkbacks.

Then somewhere Google said: you need more content to get higher ranking. Websites started to hire content writers to create extremely light, usually useless posts about problems. I think website owners could get the questions about what people were looking for so they could react to those questions.

Then, that was followed by AI generating those slop pages.

Great example of that is minitool: https://www.partitionwizard.com/news_en_sitemap.xml. There is no reason for an article like https://www.partitionwizard.com/news/cable-not-connected-on-... to appear on their website. And notice every article is ended with

> By the way, MiniTool Partition Wizard is a reliable and professional partition manager. It can convert MBR to GPT without data loss, migrate OS, clone a hard drive, recover data from a hard drive, etc. If you have this need, download it to have a try.

Anecdotally, I was searching for Jerusalem bugs after having caught on on a hike. First result that popped up was from a pest control stating their bite causes fever. No, they don't, but it's a great way to attract traffic (talking about bugs!) and then scaring people into buying their services. It's extremely difficult to consider what they write to be the authority as they aren't neutral in this transaction.

Then all the AI models themselves started to extract 'answers' from those websites and present them to you as authority.

There are people who upload fake videos to YouTube with the most common 'KB1234567 FIX HERE' to shill their products. (Microsoft isn't helping here with their independent advisors always suggesting dism and sfc).

The websites have no incentive anymore of providing you short-form actual answers. It's gone. You can minimize it by using search engines that don't provide you with AI answers, and learn to spot AI content and move on as fast as possible. If the answer is not in the first paragraph you won't find it.

Companies are already talking about that. At least, I've been to a few conferences where SAIO (I forgot what they called it) was mentioned and elaborated upon.

I vaguely remember that it was elaborated upon to build a whole suite of tools that could predict what the chances are - given the content that you have - that you too would rank in the AI overview.

Ugh. I like websites for humans. I'm firmly opposed to having all my knowledge pre-digested by a third party.

SAIO = Search Artificial Intelligence Optimization[1]

I have a website that has been receiving traffic from chatgpt.com but the prompts users are entering for my website to appear as a reference / link aren't made available so how much optimization can really be done?

[1] https://datascientest.com/en/all-about-saio

Reworded: We've heard from sundar that users want AI responses for more of their searches

There are probably pieces of code that are dead but nobody knows they’re dead and the reason they’re dead is not discoverable by static analysis. Like junk DNA, just relics of a bygone past being carried needlessly into the future…

Google were the first company to release a product called Deep Research, which is very similar to the product OpenAI launched later: https://blog.google/products/gemini/google-gemini-deep-resea...

Horrific take, google search is already bad enough. Defaulting to chatgpt style would make it useless. GPT is not a search engine.