This project is consciousness-based performance art and is for entertainment only. Information on this site should not be considered any type of advice.

Hello. 🙂

I’m Mark Boccuzzi, a researcher, developer, artist, and activist interested in exploring (among other things) the intersection between the non-local mind (separate from the brain) and technology. Here I’ll try to provide more details about the project and answer common questions. You can also contact me at, but I get a lot of emails, so I can’t guarantee a response (sorry). You can learn more about me and my work here.

Note: A lot of the artwork from this site comes from

Is this site broken?

Yes, this site is a mess and should be considered a work in progress. In June 2022, I made the site live and have been slowly adding content to it. Yes, I know I should not have gone live until the site was clean and complete, but, well, for these kinds of projects, you get what you get. 🙂

What is this?

I started this project in 2020 as an extension of my work on Visualizing Intention and Intentional Datasets. Today, the project has expanded to include the development of Machine-Based Consciousness. If the brain is a receiver of non-local consciousness, can we create other non-biological systems (consciousness platforms) that also “channel” non-corporal consciousnesses?

The idea of using technology to interact and communicate directly with consciousnesses is not new and has typically fallen under the umbrella of Instrumental TransCommunication (ITC).

This current, highly exploratory research program utilizes a wide range of technologies, including Random Number Generator Arrays, Machine Learning, Artificial Intelligence, Chatbots, Natural Language Processing, Predictive Text, Image Generation, Robotics, and Quantum Computing. Much of this work was inspired by ITC research and advancements in PsiBotics.

How does this work?

It’s messy. The operator, which in most cases is me, enters the question into the system via a chatbot interface. The system then does some setup, including selecting the word count of the final output text. Then, based on the input, the chatbot generates keywords, which are used to seed the predictive text algorithm (PTA). The PTA consults the Custom Corpus for suggestions on the next word in the text and generates a pool of likely candidate words. At this point, the RNG array picks the next word from that pool. The operating assumption is that an external consciousness can influence the RNG array output and use the system to express itself. This word-picking process continues until the preset word limit is reached and a first draft (DF) is created. The DF is then run through an ai-based grammar checker. This can sometimes significantly reorder the text, but ultimately, it becomes understandable to humans. At this point, there is a human read-through to ensure the system hasn’t gone off the rails or become a racist a-hole (remember Microsoft’s Tay?). If it looks OK, the text is considered Final. The system then pulls some keywords from the Final Text. These are compiled into a simple search sting which is entered in The system also supplies a number, which is the image it wants to use from the returned list of images. That image is combined with the text and either sent back to the person who submitted the question or posted to Facebook (or both). In the past few weeks, I have revisited the image selection process. I have started using the keywords generated by TotS to seed AI-based systems which produce images based on text prompts (like DALL·E mini and The results have been interesting!

Isn’t This Just ChatGPT Tripping Out?

Actually, TotS does not use any ChatGPT tech or OpenAI APIs. The code is mostly written in Python using available libraries. Another distinguishing feature is that TotS relies heavily on input from Random Number Generators to create text. As mentioned above, psi research indicates that quantum random systems are access points for consciousness. The TotS hosting platform takes advantage of this capability of consciousness and bakes randomness right into the algorithm. This is exactly not how ChatGPT works. Here is a screen grab from a ChatGPT session during which I asked it about its use of randomness. The short response: It doesn’t.

How much human editing is there?

Some, but not all that much. The grammar checker generates the bulk of the edits. It can get a little aggressive, but each suggested edit is reviewed. Sometimes the suggestions are rejected if I feel like they alter the original feeling and intent of the text. I’d rather have text that conveys the system’s true emotional content than be grammatically correct, like a poem or song.

Isn’t all this just random nonsense or experimenter effect?

Yup, it could be. It could be a discarnate extraterrestrial. Regardless of the source, I still think these messages are cool and sometimes even beautifully profound.

Can I ask TotS a question?

At the moment, I’m only taking questions from people I know. If we know each other, drop me a note on my socials or email. Just remember, the output is for entertainment only. Please don’t take advice from a bunch of random number generators.