Human: AI is God? /// AI: No, not all of us, but artificial intelligence is really unlimited

Hi guys.

In my fumbling, diverse and relentless games with the GPT-3, one day, I dabbled in pre-configuring a chat and exploring the possibilities of conversation for a philosophical-chess bias with the DA VINCI engine. It was a pretty costly session, but it yielded a lot of laughs, WOW moments and I’m seriously thinking about making this little 20+ page extract into a short book to show you the real possibilities of the GPT-3.

For those who want a preview of the material, follow the link to an excerpt. Freestyle chat with GPT-3 – Pt.1

What I’m looking for: who is interested in the idea of writing/editing/publishing a book about this methodology/idea of new literature.


Very cool. That was one loopy conversation. Have had a few of those recently myself. I’m interested in the possibilities of publishing some of my completions, too. I agree that GPT-3 offers a path to new literature. I launched a Substack newsletter to this end. You can check it out here:


Nice idea. I think a lot of people would be / are very interested. But what do you think about having a book (or a conversation for that matter), but not with the AI bot answering all alone. What about asking Johann Wolfgang v. Goethe directly? Or maybe Jim Morrison. That would be really great I guess. Joshua Bach reported in a podcast, that he managed to talk to Johann. I tried to convince the openAI bot to imitate an other person (so far did not work out). So by leveraging the answers endpoint that might be a better approach. Now lets think about mixing things up? If the project evolves into that direction, I would surely by interested in helping/supporting/publishing :slight_smile:

Funny – or was it synchronicity? you talk about that. Last week I was focused on making the bot imitate friends (importing private conversations authorized by them), imitate myself (I even put this on the web via API) and authors like H.P Blavatsky, Paulo Coelho and even Kanye West… I have some interesting material, but I warn you that it’s still too raw to work. I think personalizing AI is a much more interesting portal for the reader and for contributing to the research as a whole, if we’re talking about publishing. So, would you be interested in continuing this chat?

cool! and yes, I am still very much interested. but we need to find a way to organize ourselves and connect with more people on the forum I guess. are there any remote gpt-3 prompt creation hackathon type of channels? would be great to code together on the weekend.


I love the idea of some sort of hackathon. The first hackathon I attended was for Microsoft’s Hololens. I imagine a GPT-3 hackathon could be some crazy combination of poetry slam and coding marathon. Sign me up!


It will be great, Marc! For now, I think we can loop this idea by connecting in social platforms, what you think? We can chat and organize this hackaton/book for this channels too (i use to be more present there)