The principles of training and creating your own AI

So far, this chapter has mostly consisted of a lot of services you can look at and pay for to take advantage of some of the forward-looking ideas this chapter discussed. Throughout this book, I have tried to present options that give you as much flexibility and freedom as possible, preferably open source, free, and that give you the option to build upon them sustainably and stably. The current wave of AI tools has a lot of issues, which I cover throughout this chapter, but relevant to the current discussion is that the data sources of LLM-based AI tools are fundamentally different from what you might be used to. The code behind an AI tool might be open source, but this doesn’t necessarily mean the model that powers it is. The LLM data model behind an open source tool isn’t quite like a database you can potentially dig into and look at. Unless it’s connected to an application, you have specialized tools, or the...