This web app uses cookies to compile statistic information of our users visits. By continuing to browse the site you are agreeing to our use of cookies. If you wish you may change your preference or read about cookies

How Vizologi uses A.I.


Intro


Vizologi is powered by OpenAI's GPT-4 Turbo, the world's most sophisticated LLM (Large Language Model).

GPT-4 Turbo is a neural network of approximately 1.8 trillion parameters across 120 layers, making it one of the largest neural networks in history. It uses a Mixture Of Experts (MoE). GPT-4 Turbo utilizes 16 experts within their model, each with ~111B parameters for MLP.

What does it include? The training data included CommonCrawl & RefinedWeb, totaling 13T tokens; it has the entire Internet and an extensive collection of textbooks, along with our proprietary data about companies' business models, which has been trained and specialized in our specific use cases.


AI-generated text


Once GPT-4 Turbo is fed a prompt, it generates text streams by predicting the possibility of a sentence existing in this world.

Query:

‘cats’ and ‘social media'

AI-generated ideas:

  • Catspiracy.
  • Cats are taking over the world. But what if they were actually running it all in the shadows? This is the story of the rise of cats to global dominance, starting from the end of the 20th century.
  • A cat-based dating app.
  • If Tinder can create a dating app based on proximity, why can’t we create one based entirely on cats?
  • A cat dating site.
  • Cats are popular. They’re probably hated, but they’re still popular. So, why not make a dating site dedicated to people who like cats?

One idea is to do text generation. GPT-4 Turbo learns to generate natural language by predicting the next word in an article from the previous words.

where are we going where are we going

By predicting what word comes next in a sentence, the AI must learn how to use other words in its context. This implicitly forces GPT-4 Turbo to learn many other essential general knowledge.


Strategic insights network


strategic insights network strategic insights network

I. Content acquistion

We crawl millions of different sources every day, from financial news and business information to company blogs.

II. Information extraction

Our proprietary AI models identify various entities (sectors, categories, business patterns, and companies) and relationships between them.

III. Organization identification

Organization entities are linked to unique IDs (with domains) in our database for further manipulations.

IV. Manual verification

We employ data analysts who monitor and verify data on a daily basis to ensure the data is of highest quality.