#aioverviews — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #aioverviews, aggregated by home.social.
-
Wikipédia perdeu acessos após resumos de IA do Google, diz estudo
-
The Web marketing community never learns from its mistakes. I'm beginning see huge numbers of obviously spammy YouTube videos crowd into certain types of queries on Google. This garbage content is obviously there only to influence the AI Overviews. Every time y'all come up with one of these genius ideas and you flood the Web with low-quality garbage, you end up crying and moaning that Google Done You Wrong when they finally crack down on it. You should really learn how to create real value. You won't have to flood the Web with garbage.
#seo #searchengineoptimization #youtube #ai #aioverviews #google #webmarketing #spam
-
With the way many SEO agencies appear to be misrepresenting what AI Overviews are, how they are produced, it seems like it's only a matter of time before they convince some client to do something really stupid. AI Overviews start with what is in the search results.
They are summaries of the information the search engine has indexed. Yes, sometimes the models hallucinate and make up stuff. But most hallucinations come from conflating opposing or contradictory information and opinions found in the search results.
Parody and satire in the sources have led to many a fascinating and sometimes humorous AI generation. One can argue that it's on the search engines to figure out which sources to trust. But they DO make that attempt (I'm not saying it's enough).
And that leads to yet more misinformation coming out of the SEO world. Simply flooding the index with listicles isn't enough to ensure that an AI summary will say what you want it to say.
What we've been telling our premium newsletter subscribers for a long time now is that they need mentions on highly reputable sites (and I'm NOT talking about "domain authority" and all that nonsense). But they also have to publish reliable, authoritative information (not simply rehashes of whatever their freelancers or AI tools find on the Web).
We now see a lot of SEO agencies offering that same advice. And that's a good thing. But the disconnect between these two positions is making it hard for people to understand what is happening in the search results.
The AI summaries ALWAYS start with what is in the search results. You don't have control over what they will say. And every search produces a relatively unique experience for the user based on their search context. You can't control that. You can't influence it.
At best, you can target a range of user search contexts to which you want your content to be relevant.
#seo #ai #bing #google #search #searchengines #searchengineoptimization #content #aioverviews #aisummaries #marketing
-
I posted this about a year ago. Might be a good time to repost it.
"Search With Stateful Chat" patent (Cf. https://patents.google.com/patent/US20240289407A1/en ) - appears to describe the Gemini app for smartphones.
"Method for Text Ranking with Pairwise Ranking Prompting" (Cf. https://patents.google.com/patent/US20250124067A1/en ) - documents an experimental process described in this research paper titled "Large Language Models are Effective Text Rankers with Pairwise Ranking Prompting" (Cf. https://arxiv.org/pdf/2306.17563 ). There is no indication this was introduced into a live agentic system like Gemini.
"User Embedding Models for Personalization of Sequence Processing Models" (Cf. https://patents.google.com/patent/WO2025102041A1/en ) - documents an experimental process for improving recommender (sub-)systems (like movie searches) that incorporate large language models. The process is described in this research paper titled "User Embedding Model for Personalized Language Prompting" (Cf. https://arxiv.org/pdf/2401.04858 ).
"Systems and methods for prompt-based query generation for diverse retrieval" (Cf. https://patents.google.com/patent/WO2024064249A1/en ) - updates a 2022 patent for a process named PROMPTAGATOR that generates queries more efficiently based on a small number of examples, as described in this research paper titled "Promptagator - Few-shot Dense Retrieval from 8 Examples" (Cf. https://arxiv.org/pdf/2209.11755 ). This could be used to generate query fan-outs (but query fan-out has been used in multiple systems at least since the 1990s, so there are many implementations).
"Instruction Fine-Tuning Machine-Learned Models Using Intermediate Reasoning Steps" (Cf. https://patents.google.com/patent/US20240256965A1/en ) - documents an older method for fine-tuning instructions submitted to LLMs, as described in this 2022 research paper titled "Scaling Instruction-Finetuned Language Models" (Cf. https://www.jmlr.org/papers/volume25/23-0870/23-0870.pdf ). The work has been superseded by this paper titled "Mixture-of-Experts Meets Instruction Tuning: A Winning Combination for Large Language Models" (Cf. https://arxiv.org/pdf/2305.14705 ).
This is the AI Overviews patent, titled "Generative summaries for search results" (Cf. https://patents.google.com/patent/US11769017B1/en )
#google #aioverviews #aimode #machinelearning #search #searchengines #generativesearch #seo #searchengineoptimization #webmarketing #digitalmarketing #ai #patents
-
ICYMI: Google gives site owners a toggle to exit AI Overviews and AI Mode: Website owners get a domain-level switch that clears content from AI features in 1-2 days, but it does not stop model training. Page control follows March 2027. https://ppc.land/google-gives-site-owners-a-toggle-to-exit-ai-overviews-and-ai-mode/ #Google #AIFeatures #DigitalMarketing #WebsiteManagement #AIOverviews
-
While everyone else is whining about AI summaries, some of us are counting the clicks. #seo #search #webmarketing #google #aioverviews #ai
https://www.seo-theory.com/your-sites-receive-more-traffic-than-you-realize/
-
For the record, I'm still going to talk about Answer Engine Optimization because there ARE answer engines out there. They use search engines (as I've been saying all along) but the first step in #AEO is to be indexed and ranking well enough in the search engines that the answer engines are using, such that they'll use your content.
Publishers have to face reality. The LLMs are trained on SOME of your content, yes, but they're getting your content for their answers from the search engines. If you want to be found in Web search, you have to be indexed in the search engines. If you want to be found in AI summaries or AI chats, you have to be indexed in the search engines.
And if you don't want to play in that sandbox, you still have social media, video platforms, and advertising to work with.
#seo #searchengines #searchengineoptimization #answerengines #ai #aisummaries #aioverviews #machinelearning #marketing
-
Your AI chatbot - no matter how well-trained - is NOT "a domain expert" or some kind of profession no matter how you word your prompts . It's an application that searches the Web and spits back probabilistically amalgamated summaries of whatever it found in the search results that seem to match your questions.
#ai #chatbots #llms #machinelearning #seo #searchengineoptimization #aioverviews #aisummaries #aichat #algorithms
-
The Large Language Models are not retrieving information from the Web. They are submitting queries to the search engines. If the search results contain bad results the LLMs =MAY= ignore those bad results or they =MAY= include them in their summaries. That's not "hallucination". And if the results change because you use a different LLM or revise YOUR query, again, that's not the LLM being ignorant or hallucinating - that's just the LLM summarizing the results it was given by the search engine.
Most of the problems with AI Overviews are coming from the search results because Google has the largest index of all the major search engines and it's very tolerant of multiple points of view.
#searchengines #google #llms #machinelearning #aioverviews #seo #searchengineoptimization #aisummaries
-
Sharing again. This is the most recent article on SEO Theory.
"AI Summaries Have Created 2 Unsolvable SEO Problems"
#ai #aisummaries #searchengineoptimization #seo #webmarketing #google #bing #aioverviews
https://www.seo-theory.com/ai-summaries-have-created-2-unsolvable-seo-problems/
-
#PenskeMedia, owner of #RollingStone and #Billboard, sued #Google, alleging its #AIsummaries use their journalism #withoutconsent and reduce website #traffic. The lawsuit claims Google leverages its search dominance to impose terms on publishers, impacting advertising and subscription revenue. Google argues that #AIoverviews benefit users and drive traffic to a wider variety of websites. https://www.reuters.com/sustainability/boards-policy-regulation/rolling-stone-billboard-owner-penske-sues-google-over-ai-overviews-2025-09-14/?eicker.news #tech #media #news
-
Hitman on iOS, martial arts survival and other new indie games worth checking out
Welcome to our latest recap of what’s going on in the indie game space. One very well-known indie…
#NewsBeep #News #Technology #AIOverviews #GB #Hitmangames #HitmanWorldofAssassination #SweetCarole #UK #UnitedKingdom #upcominggames
https://www.newsbeep.com/uk/104402/ -
"Search With Stateful Chat" patent (Cf. https://patents.google.com/patent/US20240289407A1/en ) - appears to describe the Gemini app for smartphones.
"Method for Text Ranking with Pairwise Ranking Prompting" (Cf. https://patents.google.com/patent/US20250124067A1/en ) - documents an experimental process described in this research paper titled "Large Language Models are Effective Text Rankers with Pairwise Ranking Prompting" (Cf. https://arxiv.org/pdf/2306.17563 ). There is no indication this was introduced into a live agentic system like Gemini.
"User Embedding Models for Personalization of Sequence Processing Models" (Cf. https://patents.google.com/patent/WO2025102041A1/en ) - documents an experimental process for improving recommender (sub-)systems (like movie searches) that incorporate large language models. The process is described in this research paper titled "User Embedding Model for Personalized Language Prompting" (Cf. https://arxiv.org/pdf/2401.04858 ).
"Systems and methods for prompt-based query generation for diverse retrieval" (Cf. https://patents.google.com/patent/WO2024064249A1/en ) - updates a 2022 patent for a process named PROMPTAGATOR that generates queries more efficiently based on a small number of examples, as described in this research paper titled "Promptagator - Few-shot Dense Retrieval from 8 Examples" (Cf. https://arxiv.org/pdf/2209.11755 ). This could be used to generate query fan-outs (but query fan-out has been used in multiple systems at least since the 1990s, so there are many implementations).
"Instruction Fine-Tuning Machine-Learned Models Using Intermediate Reasoning Steps" (Cf. https://patents.google.com/patent/US20240256965A1/en ) - documents an older method for fine-tuning instructions submitted to LLMs, as described in this 2022 research paper titled "Scaling Instruction-Finetuned Language Models" (Cf. https://www.jmlr.org/papers/volume25/23-0870/23-0870.pdf ). The work has been superseded by this paper titled "Mixture-of-Experts Meets Instruction Tuning: A Winning Combination for Large Language Models" (Cf. https://arxiv.org/pdf/2305.14705 ).
This is the AI Overviews patent, titled "Generative summaries for search results" (Cf. https://patents.google.com/patent/US11769017B1/en )
#google #aioverviews #aimode #machinelearning #search #searchengines #generativesearch #seo #searchengineoptimization #webmarketing #digitalmarketing #ai #patents