I have purposely avoided publishing any of my writings on the internet related to my project: The Human DataBit. Rather, I have chosen to publish meta-content generated by LLM-based AIs, trained from my original material. Mostly to avoid the insane web scraping that is feeding the LLM rush (see my post-https://www.linkedin.com/pulse/end-internet-toolifyai-ai-content-farms ). Second, to prove the point of Ambient General Intelligence & The Human DataBit (HDB), we need AI to prove what is and, what is not generated by AI.

I uploaded 40-50 pages of my notes on The HDB to Google’s NotebookLLM and used the deep dive conversation function to check out the feature. I must admit I was surprised by the output of the LLM, the conversation is weirdly interesting, and slightly engaging, and is the AI version of bad hype content. Better ways to sell you, Shaklee.

How much meta-content constitutes real content

It is bizarre to hear a two-way conversation about my ideas. Having an algorithm collate and organize a perspective and analysis of my writing and then discuss it, takes some getting used to it at first. It is sort of like the painter who turns the painting upside down and steps back to gain a new perspective, I guess. I did find a few ideas for how I can be clearer in my content.

As for the output by Google, it is cute, shallow, and syrupy in delivery with a bunch of circular and reductive statements, yet an impressive feat of technology. More important to me, is the input by me into the pantheon of data used for AI training. As I give new info to the LLMs from the original content, there is a meta-training that happens as a function of the “exchange”.

Overall, I am intrigued by the podcast content. It does a good basic job of explaining some of the content I uploaded. It has some value in planting the seeds into the data stream. I created a Soundcloud page and uploaded it as an introduction to The Human DataBit, along with other centralized systems. I watch the crawlers hit the website in the hours after posting anything to the data broker sites (LinkedIn, IG, FB, X, etc.),

The sum of most knowledge and content created before the current AI boom has been incorporated into the “training” of the models we now interact with as LLMs. Each interaction creates new paths of training information available to the AIs. AIs & Algorithms feed us content and each interaction we have with it creates new training data for the AIs/A’s, to feed us more content. Ruled by reductive (predictive) logic.

“In all chaos, there is a cosmos, in all disorder a secret order.” ― Carl Gustav Jung

The fabrication capacity of current AI models should serve as an alarm for everyone, awareness of this “brave new world” is an adaptive strategy. Being an informed consumer of content is now more important than at any other time in our history.

AI needs data. AGI needs even more data. As a species what differentiates artificial intelligence from biological intelligence is our ability to generate data the AI can’t. The ability to generate Curated & Quantified personal data is the new economy.

AI is ubiquitous and pervasive, AI driven Personal Data Strategies are a new territory and a new social-value, data-value economy. An economy based on using AI to prove what is, and is not created by biological intelligence. The value of data provenance and clear attribution for human personal data should be a basic right.

Quantity over quality/accuracy of AI remains to be improved and remember Artificial Super Intelligence is on the horizon.  For more details, please listen to the NotbookLLM deep dive conversation on Soundcloud. It provides an interesting perspective of The Human DataBit.

Link to SoundCloud

LinkedIn