Welcome to the Layer Cake

“Welcome to the layer cake, son.” Mr. Temple utters these words to a pre-James Bond Daniel Craig as he explains the “Facts of Life” in the 2004 film Layer Cake. 

The Generative AI ecosystem can also be described as a Layer Cake, with (at least) three layers: 1. Foundation Layer 2. Middle Layer 3. Application Layer.  

The Foundation Layer consists of large, complex, and hard to develop Large Language Models (LLMs) such as OpenAI’s GPT-3.5. These foundation models are general models, trained to be able to generate responses across a wide variety of subjects.  

The Middle Layer takes the models of the Foundation Layer and improves them, creating more specialized models that are better at solving particular problems, serving particular verticals, or enabling the work of a particular organization.

The Application Layer takes a model and makes it usable and useful to an end user to solve their problems.  

Where is boodleGPT in the Layer Cake?

boodleGPT is part of the Middle Layer of AI. We’re taking a Foundation model (OpenAI’s GPT-3.5 or GPT-4) and optimizing it with 35 billion proprietary predictive insights to create a Middle Layer model that does something OpenAI’s GPT can’t do on its own: generate predictions about the behavior of 250 million adult Americans. We will then build additional Middle Layer models that are focused on particular verticals (philanthropy and business) and/or customized to generate responses for particular organizations using their own data.

Our first application layer platform will be a chatbot powered by boodleGPT, but we will quickly follow with a plugin that allows other platforms to integrate boodleGPT as well as an API that allows other applications to be built using boodleGPT.

In the next few posts, we’ll talk about each of these layers, our path forward (and reasoning), and what it means for our future users and partners.  

Welcome to the Layer Cake, it’s gonna be delicious.

.

_{References:
Dr. Andre Retterath | LinkedIn
Data-driven VC #19: Value accrual in the modern AI stack (datadrivenvc.io)
AI for the Next Era | Greylock
Generative AI: How To Spot Real From The Hype | Moveworks
ChatGPT, LLMs, and Foundation models — a closer look into the hype and implications for startups | by Oliver Molander | Better Programming
The generative AI revolution has begun—how did we get here? | Ars Technica}