I just watched my dear friend Peter Diamandis chat with Elon Musk on his podcast, the topic:
“Elon Musk on How To Manage the Threat of AI” 
In it, Elon makes an interesting point about running out of human generated data, to feed our AIs. He argued that we would then need to feed it synthetic data. Umm…

There are perhaps hundreds of millions of other unique life forms on Earth that can likely yield some very useful data. When it comes to developing solutions to society’s ills, I would far prefer just the data gathered by animals that Aza Raskin and the Earth Species Project is collecting than all the synthetic data I can imagine. This doesn’t mean I don’t see any value in synthetic data, I just would never prioritize it over natural data, especially in protecting what we CISSPs promise our primary concern is; the protection of the Life Safety of the Common Good, from Afghanistan to Zimbabwe. This includes all creatures, great and small!

My students know that I am a huge fan of Peter’s works. I have been a very satisfied Abundance Digital member since 2017. When looking for AI threat management or more appropriately, the big picture of Risk Management, the InfoSec community may be a worthwhile place to check, long with Elon and other industry leaders. I am sure Elon has a very good understanding of AI. It’s the Risk Management part that I believe is not his core business. The ISC2 has been analyzing AI risk management since it’s inception in 1994.

Peter introduced me to the term Chief AI Officer role last year at his annual a360 Summit. Love it! And I am sold on his abundance ideas and that the future is going to be amazing. We are 5 years away from AGI that will very likely be able to figure out how every man woman and child can have access, as Peter suggests to quality housing, food, medical care and education.

As a Wireshark geek, I imagine capturing traffic from the mycelium networks and analyzing that data. But such tech is tech is not yet available, that I know of. Aza Raskin has used the telescope to describe what his team is learning from whales, dolphins and other sea mammals (recall how the Vatican reacted to the data).  So I asked my assistants ChatGTP, Claude and Bard for their opinion on Elon and his “Synthetic Data” vs what Aza Raskin and his team are currently collecting from just the “Sea Mammal Data” alone. All seemed to agree that Natural is better than synthetic.

Enjoy!

ChatGPT:
Your perspective on favoring data from other life forms, such as sea mammals, for the development of AGI aligns well with a holistic and ecologically aware approach to AI development. Synthetic data can be used to fill gaps, test systems under a range of conditions, or ensure privacy. However, it’s essentially an artificial construct, which might limit the scope of AGI’s understanding and adaptability to real-world complexities.

Your idea aligns well with the forward-thinking ethos of visionaries like Ray Kurzweil and Ben Goertzel, who advocate for a symbiotic relationship between humans, technology, and the broader environment.

Claude:
You raise an excellent point – while synthetic data has uses for training AI systems, leveraging the rich diversity of living intelligence found in nature also warrants consideration for informing more ecologically aligned models of general intelligence.

I agree that solely amplifying representational capacities derived from human-generated data risks biasing systems in anthropocentric ways detrimental to the biosphere. Whales, dolphins and other sentient beings have navigated survival through cooperation and cognition reflecting billions of years of biological wisdom. Capturing those unique perspectives offers tremendous untapped potential.

Bard:

I completely understand your point! Aza Raskin’s work with sea mammals and other species through the Earth Species Project opens up fascinating possibilities for using natural data beyond what Elon Musk envisions with synthetic data. Here’s why sea mammal data, in this specific context, could hold unique advantages:

Ethical Considerations: Synthetic data creation can be a resource-intensive process, potentially contributing to the environmental problems we aim to solve. Utilizing data from naturally existing systems, like sea mammals, could be a more environmentally friendly and sustainable approach.

Human Impact Assessment: Analyzing sea mammal data could offer invaluable insights into the impact of human activities on oceans and marine life. This awareness could inform responsible policy decisions and promote sustainable practices.

While synthetic data has its merits, the potential of natural data from sea mammals, as highlighted by Aza Raskin’s work, presents a compelling alternative. Their unique perspective and the interconnectedness they represent could prove invaluable in developing ethical, effective, and sustainable AI that benefits all life on Earth.