Recently Tom Vassel released a video, showing off Google’s new AI, NotebookLM. Unlike ChatGPT, which just scrubs the internet, this one has you upload a PDF to take data from, and this is a very interesting approach, as it gives the user a great deal more control over how the system works.

When I released The Nullam Project, I also released an accompanying PDF filled with prompts that you could feed into ChatGPT to teach the AI, so it could generate characters and make missions, but it had it’s flaws. To say that ChatGPT was confused is an understatement, and because it is based on popular content, when I asked to build a pregen character for The Nullam Project, I would often get stats for 5e instead.

With NotebookLM, I uploaded the PDF for The Nullam Project, at which point the AI gave me a very accurate summery of the game, but the really impressive part is when I asked it to build a character for a one shot, …and it did it flawlessly. Not just getting the rules of the game right, but actually picking stats, skills and equipment that made sense for the class it selected.

So, will I be using this for my designs? …no, but I will be telling people to upload my games into it and see how things work. It doesn’t do much for me, aside from the novelty, but it has the potential to be a very powerful tool for my customers, which indirectly benefits me.

The other question is, do I fear that this thing will replace me as a developer? …also no. People like human made things. Especially if there is a physical version, and that means that I will be here for years to come, but I do acknowledge that AI is inevitable. I don’t use it for any of my design stuff, but I understand that there will come a day when AI becomes the norm, which hopefully is presented as a tool to speed things up, rather then the current, enter a prompt, get an image system we have now.

For now, I just suggest that you upload The Nullam Project and see how the AI works. Even if you are opposed to AI, you need to to be aware of how it is progressing and what it is capable of, because the worst argument you can have is one where you have not studied your enemy, and in this case, it’s free and open to everyone.