Google’s AI Merger, Project Magi, and the Rise of Open Source Language Models


Hey there! It’s TJ from Envision, and today we’re going to discuss the most recent SEO and AI news, focusing on how it applies to local service businesses like our clients. Although this update is mainly for our internal team, we thought it would be valuable to share it with everyone. Let’s dive into the exciting world of AI and SEO!

Google Merges Brain and DeepMind

In a major move, Google has decided to merge its two biggest AI departments, Brain and DeepMind. Previously operating as separate entities, this merger could be a response to growing competition from Microsoft. The new combined entity’s name is still unclear, but the move hints at Google’s commitment to advancing its AI capabilities. As the AI wars continue, we’re eager to see what innovations come from this collaboration.

Google Magi and the New Search Engine

There’s some confusion around Google’s Project Magi, but one thing is clear: exciting changes are on the horizon. According to Barry Schwartz, Project Magi refers to new features coming to Google’s existing search engine. On the other hand, the New York Times claims that Project Magi is actually a new, AI-driven search engine with a personalized, conversational interface. Regardless of the specifics, we can expect significant advancements in search engine technology from Google.

Transactions Built into Search

One of the primary goals for Google is to keep users on its search engine results page for as long as possible. To achieve this, the tech giant aims to enable transactions directly on its platform. While this may not be ideal for website publishers, it’s a trend worth watching. As Google makes it easier for users to buy products and book services without visiting individual websites, it’s crucial for local service businesses like ours to adapt and ensure we maintain visibility.

Stability AI Launches StableLM – Its Own Language Model

In the world of open-source AI, Stability AI has launched its own large language model called StableLM. While not as big as some other models (it currently has 3 billion and 7 billion parameter models), StableLM is noteworthy because it’s open source. This allows developers to take the model and use it for various applications, potentially leading to the creation of highly specialized AI tools.

Sliding Window Processing for GPT-4

A recent tweet from an OpenAI team member mentioned the upcoming release of a sliding window-like processing feature for GPT-4. This update will allow the AI to process and summarize book-sized documents, expanding its capabilities significantly. Once this feature is available, there could be numerous applications for content generation and data analysis. We’re looking forward to seeing how this improvement impacts the AI and SEO industries.

Generative AI Coming Soon to Google Ads

Generative AI is making its way into Google Ads, with the tech giant experimenting with AI in their ad editor for responsive search ads. This could result in Google recommending ad copy using technologies like BERT or Lambda. As this feature is built into the platform and uses data from your account, it could prove to be highly beneficial.

Canva’s AI is Getting Pretty Crazy

Canva, a popular graphic design platform, is pushing the boundaries of AI integration. One recent example is a feature that allows users to seamlessly change an image’s elements, like the color of a blouse, with just a few clicks. Though this may not be directly useful for SEO professionals, it demonstrates the rapid advancements in AI and its potential applications.

Google Removing Some GSC Reporting in December

Google has announced plans to remove some Google Search Console reports, specifically page experience reports, starting in December. This includes the mobile usability report. While it’s disappointing to lose access to this data, Google’s decision might be influenced by the overemphasis on these reports. Google wants users to focus on content and user experience, rather than metrics like page load time or cumulative layout shift.

Google Introduces Google Reader Revenue Manager

Google is introducing the Google Reader Revenue Manager as a way to compete with platforms like Substack and Twitter in the creator economy. This tool allows publishers to easily create paywalls and subscriptions for their content. It enables users to subscribe to content through the Google Play Store, and makes it easier for creators to monetize their content without relying on third-party platforms.

Google Showing FAQ Results Much Less Often in Mobile

Google has reduced the frequency of FAQ results in mobile search, although there hasn’t been a significant change in desktop search. Despite this, it’s still recommended to add FAQ structured data to your pages, as it can help Google understand your content better.

Cool Tools/Techniques:

MiniGPT-4 (Bimodal ChatGPT)

One interesting tool is MiniGPT-4, a research project using GPT-4 that can take in images and generate relevant text based on the image. This feature will likely be integrated into ChatGPT soon, but in the meantime, MiniGPT-4 is a free tool that can be used to experiment with image-to-text AI.

Animated Drawings

Meta has released a fascinating tool that allows you to animate drawings by simply uploading a picture. While not directly applicable to the SEO world, it’s a fun and creative way to engage with AI technology, especially if you have kids.

As AI technology continues to advance, it’s essential to stay up-to-date on the latest developments and find new ways to utilize these tools in your work. Stay tuned for more updates and insights in the world of SEO and AI.


As AI and SEO continue to evolve, it’s essential for local service businesses like ours to stay informed and adapt to the changing landscape. From Google’s merger of Brain and DeepMind to the potential of open-source language models like StableLM, we’re excited to see what the future holds. Stay tuned for more updates on the latest SEO and AI news and how it impacts local service businesses!