The assistant on your phoneis doing something you can watch.
Part 1 ended on the next-word trick. This part opens it up: how words become numbers, what the model is looking at when it answers, why it forgets, how it reads your documents, when it invents, and what happens when you let it act. Every screen is a small working model.
Six ideas between the next-word trick and an assistant you would trust
You use one of these every day. You know it is fluent, you suspect it is sometimes wrong, and you have been told it can be connected to your systems. This part shows the machinery behind each of those feelings, so that “we just plug the AI into our data” becomes a sentence you can picture.
On the screen. The loop behind this page: words flow into a network and come out as a sentence. Press begin when you are ready.
Where you have met this. By the last screen you will be able to sort any AI request into four buckets: it knows this, it needs our documents, it needs a tool, or it should not be doing this alone.
Every model on this page is a deliberately small illustration running in your browser, not a production language model. The behaviours are real; the scale is not. This is an illustrative walk-through, not client results.
The next move is a conversation
Make your next move a better one.
Bring the workflow that slows you down, the decision on your desk, or the idea you want to put to work. We will help you find a sensible next step.
A specific challenge. An honest conversation. A clear next step.