Welcome to another AI Roundup and Rundown. This summer has hosted a bit of a heatwave, and the AI news from the past month is equally hot. Here are some of the most interesting AI in marketing developments from throughout July.
Generative Engine Optimization Doesn’t Matter On Google?
Gary Illyes, a Search analyst for Google, recently said that AI SEO, generative engine optimization (GEO) and other specialized techniques to improve your chances of showing up in Google’s AI Search products aren’t necessary.
Kenichi Suzuki, a Google product expert, shared a snippet of Gary’s statement from Google’s Search Central Deep Dive event this month on LinkedIn:
“To get your content to appear in AI Overview, simply use normal SEO practices. You don’t need GEO, LLMO or anything else.”
And while it’s certainly true that solid, traditional SEO is essential for ranking on Google in any capacity or within any Search product, Google isn’t the only player in the AI search space. Other AI engines like ChatGPT rack up millions of searches daily, and emerging GEO tactics can still influence visibility in other tools.
Similar to how robots.txt tells search engine crawlers (like Googlebot) what pages they can or can’t index, LLMs.txt is an emerging file type intended to guide generative AI models on how to interpret or prioritize content in their responses. Analogous to the former, LLMs.txt isn’t a replacement, but players like OpenAI are using it right now to crawl websites, while Google is not and doesn’t plan to, according to Illyes.
All things considered, we’re taking this with a grain of salt. The way AI advancement and competition unfold changes constantly, and today’s truth could very well be tomorrow’s “we’ve adapted.”
There are two takeaways here: strong SEO still matters. And GEO increasingly matters, too — maybe just not as much on Google right now.
Your AI Assistant Is Getting An AI Assistant
On July 17th, 2025, OpenAI posted on X introducing a new tool called ChatGPT agent: “ChatGPT can now do work for you using its own computer,” they wrote.
Yes, your computer now has its own computer to help you complete even more complex tasks. In the official product release announcement from OpenAI, the company says this new “unified agentic system” can complete tasks from start to finish, like creating, sizing and ordering custom stickers that get shipped straight to your door, all from one prompt. Or, creating complete reports or decks that reflect internal business statuses based on recent news.
It’ll be interesting to see where this goes and how marketers adopt it into their processes — for better or worse.
AI-Powered Wearables Could Be Getting Blingy
Meta’s Ray-Ban AI glasses have been in the AI wearable spotlight for some time, but Samsung is interested in entering that market with all-new ideas. Chief Operating Officer for the company’s mobile experience division, Won-joon Choi, told CNN in early July that the company is interested in exploring “glasses, earrings, watches, rings, and sometimes a necklace.”
Are we hurling toward a future where marketers can whisper commands into their jewelry to complete tasks quickly — maybe without even being at their computer? I’m unsure, but AI earrings could become a reality.
Google Is Experimenting With a New AI-Organized SERP
Just when you thought things couldn’t change any quicker, Google announced a new, experimental SERP that’s organized by AI, called Web Guide.
This new SERP uses a custom version of Google’s Gemini AI to group search results into themed sections (e.g., safety tips, personal stories, step‑by‑step guides) rather than showing a flat list of links. They claim it to be most helpful for open-ended queries like, “how to solo travel in Japan,” where Web Guide might break results into sections like accommodations, budgeting, sights and more.
Why is this interesting? Well, traditional SEO has been largely about ranking high in a linear list of blue links. Web Guide changes this by clustering content into AI-generated topic sections. If the feature takes off, position 1 may no longer be the focal point for marketers and SEOs. Content visibility may depend on which section your page appears in and how well it satisfies that subtopic.
Web Guide also uses what Google is calling a “query fan-out” technique, which “takes your original question and explores multiple angles of it behind the scenes.” So, instead of treating searches as single queries, Google will essentially expand them into multiple related sub-queries, each exploring a different facet or interpretation of the original intent. This could shift marketers’ keyword strategies toward more semantic SEO tactics that cover topics thoroughly, including subtopics, variations and related questions, instead of matching exact search terms.
An AI Summer Shakeup
There are lots of interesting things to explore deeper here, from Google saying GEO doesn’t matter for ranking in AI Overviews to introducing an experimental feature that shakes up traditional search engine results pages as we’ve known them.
As the old saying goes: the only certainty is uncertainty.
