From SEO and UX to AI ChatBot Training

With over 20 years of experience in digital marketing and SEO, Nettpilot is now stepping into the artificial intelligence revolution.

Our expertise in digital insights forms a solid foundation for exploring the opportunities AI technology offers. In this article, we share our thoughts on how RAG technology can transform customer service and marketing, and why Nettpilot is well-positioned to guide businesses through this exciting development.

#AI

Written by Nickolass Jensen
Assisted by Aurora Høegh-Omdal, Kira-of-Nettpilot
In consultation with Google Gemini, Co-Pilot, ChatGPT4o.
References can be found at the bottom of the article.

This is my long, slightly nerdier, and more thorough version of the article: From Science Fiction to Reality: MemoRAG – The Future of Memory for AI.

RAG (Retrieval-Augmented Generation). R-A-G. Learn this. Understand what it actually means, both for you as a consumer navigating the web and digital channels, and for you who work with information retrieval, data collection, and insights.

It is with RAG that we build the knowledge base for artificial intelligence. RAG combines information retrieval with text generation.

RAG systems use specialized tools to fetch useful information from external databases. This allows language models (LLM = Large Language Models) to generate fact-based answers by using the retrieved knowledge as context.

An example can be found here, very well explained by Kira, who is our own chatbot (you can talk to her yourself via the chat in the bottom right corner):

Screenshot of a conversation with Kira-of-Nettpilot, conducted during the writing of this article. Talk to her yourself in the bottom right corner.

An interesting development of RAG is MemoRAG, which introduces a memory module to handle complex tasks with long inputs.

A Chinese research project from Peking University’s Department of Artificial Intelligence launched a very interesting research paper on MemoRAG last month, September 10.

This system builds a compact global memory across the entire database, enabling the generation of context-dependent cues. These cues effectively link the knowledge database to the precise information needed for accurate answers.

I acknowledge that this can be complex material for many of you, and this was also almost complete gibberish to me until recently. However, I have spent some time trying to find explanatory models that are somewhat comprehensible, so here it comes.

What is MemoRAG?

MemoRAG is a new technology that improves the way artificial intelligence (AI) retrieves and generates information. Think of it as a super-intelligent librarian who not only finds books for you but also remembers everything you have read before and uses this knowledge to give you even better answers.

How does it work?

MemoRAG uses two main components:

Memory Model: This is like a large, smart diary that remembers all important information from a vast amount of data. It can remember up to a million words at a time! Think of it as Jarvis from Iron Man, who always has all the information ready for Tony Stark.
Retrieval Model: This is like a traditional search engine that finds relevant documents based on what you ask. But with help from the memory model, it becomes much smarter and more precise. It is like when R2-D2 helps Luke Skywalker find the right information in Star Wars.

Standard rag vs memo rag — Standard rag vs memo RAG

Together, these two create an AI assistant that can be as helpful as Samantha from the movie “Her”, but without the complicated emotional implications!

The movie “Her” from 2013, directed by Spike Jonze, is a fascinating exploration of the relationship between humans and artificial intelligence. The plot follows Theodore Twombly (played by Joaquin Phoenix), a lonely man who works writing personal letters for others. He falls in love with an advanced AI operating system named Samantha (voiced by Scarlett Johansson). The film explores deep themes such as love, loneliness, and what it means to be human in an increasingly technological world. Samantha evolves rapidly, and the relationship between her and Theodore becomes increasingly complex and challenging. “Her” raises interesting questions about the future of human relationships and our interaction with technology. It shows both the possibilities and challenges that can arise when artificial intelligence becomes so advanced that it can simulate human emotions and relationships.

Advantages of MemoRAG

Better Understanding of Context: The memory model provides a deeper understanding of what you are asking, allowing the retrieval model to find more relevant documents. It is like when Hermione Granger uses her vast knowledge to find the right solution in Harry Potter.
Increased Accuracy: By using the memory model, the retrieval model can find documents that might otherwise be overlooked. It is like when Sherlock Holmes uses his incredible observation skills to find clues that others miss.
Efficiency: The memory model reduces the need to search through all information every time, saving time and resources. Think of it as when The Flash uses his super speed to do things faster than anyone else.
Flexibility: MemoRAG can be used with different types of retrieval and language models, making it very adaptable. It is like when Doctor Who uses his TARDIS to adapt to different time periods and situations.
Scalability: The memory model can handle large amounts of data and long contexts, making it suitable for complex tasks. It is like when The Matrix uses its vast computing power to simulate entire worlds.

Applications of MemoRAG

MemoRAG can be used in many different contexts, such as:

Question and Answer: Provide detailed and accurate answers by using the memory model to better understand the context.
Summarization: Create short and precise summaries of large amounts of information.
Creative Writing: Help generate content that is rich in context and details.

MemoRAG is like a super-intelligent assistant that combines the best of both memory and search to give you the most relevant and accurate answers.

It is a significant improvement in how AI can help us find and use information.

To the Bridge: From SEO and Good UX to AI Bot Building

Let’s then create the experiment and explore the connection between precise modeling of MemoRAG for AI assistants, and achieving top rankings in organic search results on Google, as well as how a skilled SEO specialist can be compared to a skilled AI bot builder. Understanding Precise Modeling of MemoRAG

MemoRAG is designed to improve how AI assistants retrieve and generate information by using a combination of memory and retrieval models. This requires a deep understanding of:

Context: AI must understand the context of the user’s query to retrieve the most relevant documents.
Accuracy: AI must be able to identify and retrieve accurate information that is relevant to the query.
Efficiency: AI must be able to do this quickly and with minimal resource use.

Understanding SEO and Top Rankings on Google

SEO (Search Engine Optimization) has always been about optimizing websites to rank higher in organic search results on Google. This requires a deep understanding of:

Context: SEO specialists must understand what users are searching for and which keywords they use.
Accuracy: The content on the website must be relevant and accurate to meet the user’s needs.
Efficiency: The website must be optimized for fast loading times and good user experience.

Comparison: SEO Specialist vs. AI Bot Builder

Context Understanding:
- SEO Specialist: Must understand the user’s search intent and optimize content accordingly. This includes using the right keywords, creating relevant content, and understanding the user’s needs.
- AI Bot Builder: Must understand the user’s query and the context around it to retrieve and generate relevant information. This includes training the AI model to understand complex queries and contexts.
Accuracy:
- SEO Specialist: Must ensure that the content is accurate and relevant to meet the user’s search. This involves using reliable sources and up-to-date information.
- AI Bot Builder: Must train the AI model to retrieve accurate information from large datasets and generate precise answers.
Efficiency:
- SEO Specialist: Must optimize the website for fast loading times and good user experience, which can affect Google rankings.
- AI Bot Builder: Must ensure that the AI model can retrieve and process information quickly and efficiently, improving the user experience.

At Nettpilot, we continue to move towards new horizons

Both a skilled SEO specialist and a skilled AI bot builder must have a deep understanding of context, accuracy, and efficiency to succeed. They use different tools and techniques, but the goal is the same: to deliver relevant and accurate information to the user in an efficient manner.

An SEO specialist optimizes websites to meet user searches and improve Google rankings, while an AI bot builder trains AI models to understand and respond to user queries intelligently.

Both roles require continuous learning and adaptation to new technologies and trends.

Both roles must stay updated on the latest trends and technologies in their fields.

For SEO specialists, this could be changes in Google’s algorithms or new search technologies, while for AI bot developers, it could be advancements in machine learning or new frameworks for natural language understanding.

It is interesting to note that these two fields often overlap, especially when it comes to understanding user behavior and delivering relevant content.

For example, a chatbot on a website can help improve the user experience, which in turn can have positive effects on SEO.

At Nettpilot, we have expertise in both SEO, Google SEARCH, PPC ADS, Programmatic, and AI development, and we often see how these fields can complement each other to create more effective and user-friendly digital solutions.

By considering and integrating an understanding of MemoRAG into our services, Nettpilot can potentially offer even more advanced and effective solutions to our customers and partners, strengthen our position in the market, and once again place us at the forefront of technological development in digital marketing and e-commerce.

“We’re at the frontiers now”