Page 1 of 1

Philosophical Project: An AI-powered Global Inquiry and Debate on Eugenics and GMOs

Posted: July 26th, 2024, 4:07 am
by value
I hereby share a philosophical project that I have been engaged in for the past month as part of the project šŸ¦‹ GMODebate.org: a philosopical inquiry with nature and animal protection organizations in over 250 countries in as much languages.

The focus of this philosophical inquiry project is twofold:
  1. To gain deep insights into the moral considerations surrounding eugenics.
  2. To unlock these insights publicly, facilitating an international "GMO debate".
For the project I use an advanced AI communication system that was created specifically for the project and that facilitates complex philosophical discussions in hundreds of languages with a level of nuance that has impressed even native writers from Paris, šŸ‡«šŸ‡· France:

Au fait, votre franƧais est excellent. Vous vivez en France ? (after about 5 emails back and forth, delving deeper into complex issues and questions).

So far, over a thousand philosophical discussions were initiated, some consisting of dozens of emails that delve deeper into subjects such as "marine philosophy".

Some examples are a conversation with DJ White who is a co-founder of Greenpeace USA and who now operates EarthTrust, Flipper Fund and Save The Whales International and Mark J. Palmer from Earth Island Institute, International Marine Mammal Project (IMMP) and Save Japan Dolphins.

These discussions quickly focused on the potential role of philosophy for marine protection and for example the work of philosopher John C. Lilly who was a first pioneer to use philosophy to understand šŸ¬ dolphin and šŸ³ whale language.

Image

"The feeling of weirdness came on us as the sounds of this small whale seemed more and more to be forming words in our own human language. We felt we were in the presence of Something, or Someone who was on the other side of the transparent barrier" ~ philosopher John C. Lilly

DJ White employed philosopher John C. Lilly in one of his own research projects. He wrote:

"I read Lillies then-current works in college in about ā€™70, first met Lilly in 1979 on Kauai and had a chance to speak with him on a number of things. Quite a mind. I wound up staying in touch, and in the last decades of his life he lived on Maui and the roles were a bit reversed, as he was a supporter and advisor of my own dolphin-intelligence research lab on Oahu, and would come over and visit. My lab had a far different set of ethical groundings than his - he was a bit of a psychopath - but he endorsed it as the outgrowth of his work. It was the first to show self-awareness in a nonhuman by human testing standards, among many other things, and existed from ā€™88-ā€™03."

My question: Interestingly, recent initiatives like Project CETI (2017), which aims to decode whale language using AI, seem to be starting from scratch. This gap between Lilly's work and current efforts suggests a lack of continuity in marine philosophy.

DJ White: "There is indeed a lack of continuity of original approaches, not that Lilly was ever considered mainstream. Iā€™ll note that Ken Norris was also an advisor to my lab and also visited, who was the other pole of cetacean research expertise; with both of those guys involved we may have been doing something rightā€¦ but I wouldnā€™t have wanted them to be there at the same time. In terms of abstract communication with the great whales, I have long sought to put that into practice but never have had the funding to do it. I think it could be profound. Iā€™m in touch with some of those who think AI will be the key, but I think there are crucial conceptual gaps in the projects Iā€™m familiar with now. AI alone wonā€™t get it done."

It concerns an expert insight that shows that the philosophical approach to marine research, for example when it concerns understanding dolphin and whale language, didn't get much further than the pioneering work of John C. Lilly in the '50s of the past century.

Overal, the philosophical discussions resulted in high value insights into philosophical topics around nature and animal protection, and more specifically eugenics or "anthropicentric GMO" and correlated general philosophy subjects such as the fundamental nature of existence and morality.

The organizations clearly appreciate the initiative and love to contribute. However, it is clear that the response-percentage per country differs greatly, with for example the response percentage from šŸ‡¦šŸ‡¹ Austria being almost 100%, with extensive and considerate responses, with even a doctor responding personally with an in-depth response in which he raises several questions, while in some other countries, the response percentage is lower, although some of the organizations that do respond communicate enthusiasm and willingness to discuss the subject.

So far I've conducted over 1,000 philosophical discussions in dozens of languages. A recent interaction with Jojo Mehta, the co-founder and CEO of Stop Ecocide International, resulted in a dedicated article about šŸ‡§šŸ‡· Brazil's attempt to eradicate the šŸ¦Ÿ mosquito species using GMO mosquitoes.

BBC: "The mosquito is the most dangerous animal in the world, carrying diseases that kill one million people a year. Should the insects be wiped out from Earth?"
https://gmodebate.org/ecocide/

Some users on this forum (the below response is from a philosophy teacher) also argued that the mosquito should be eradicated from Earth, showing that the subject is important from the perspective of the mosquito species.
Terrapin Station wrote: ā†‘February 16th, 2020, 11:27 am
value wrote: ā†‘February 10th, 2020, 8:04 am If declines continue at this pace, many insect species could vanish forever within a few decades.
Please let that include mosquitoes and roaches. And can't we just do it now instead of having to wait a few decades?
The article makes a case for the vital importance of the mosquito species and delves into pioneering topics such as "the fundamental role of šŸ¦  microbes in species relative health" and "driver of animal evolution", which is interesting from a philosophical perspective.

The šŸ¦Ÿ Mosquito: Critical For Ecosystems And Evolution
https://gmodebate.org/ecocide/

Artificial Intelligence LLM Model

The AI LLM model used is Anthropic Sonnet 3.5 AI. The quality is amazing and the AI really is able to understand implications of deeper logic when given the right prompt, and to translate that understanding into high quality texts in any language, not just grammarly correct but also strategically written for a purpose, in this case inspiring specific receipients to reveal deeper insights into their moral intuitions and considerations on the topic of eugenics and GMO.

Anthropic isn't cheap, but as mentioned in my other topic about šŸ‡¬ Google's intentional deception and 'low quality' responses with its Gemini AI (Google, known for its Do Not Evil founding principle, apparently turned 'bad guy'), one might say today that 'quality has a price' (for now, because with human labor becoming fundamentally obsolete in the near future it might be at question whether humanity should be charged for the use of good quality AI, but that would be another topic, partially discussed in topic The AI Revolution: An Utopia for Philosophy?).

Overal the project costed several thousands of USD. To complete the inquiry in the next years with hundreds of thousands of organizations, it might cost a bit more, but that would depend on the price of AI which could change in even weeks time from now.

More info about the project: https://gmodebate.org/

Philosophy eBooks: https://gmodebate.org/book/

Hopefully the project inspires others for innovative usages of AI for the advancement of philosophy.

Re: Philosophical Project: An AI-powered Global Inquiry and Debate on Eugenics and GMOs

Posted: July 27th, 2024, 7:37 am
by value
Some example questions and responses.

Questions to DJ White, co-founder of Greenpeace USA:

The founders of Greenpeace in 1971 on an island in the Arctic with a campaign to stop a nuclear bomb.
The founders of Greenpeace in 1971 on an island in the Arctic with a campaign to stop a nuclear bomb.
founders-greenpeace.jpg (57.06 KiB) Viewed 747 times

Your focus on "large existential threats to conscious life on earth" is deeply compelling. How do you see the role of philosophy in addressing these threats? Could a renewed emphasis on philosophical inquiry in marine conservation help to refocus efforts away from "techno-futures which will never exist" and towards "the profound realities of consciousness and abstract communication"?

I think philosophy will mainly be important in causing a reltively small number of humans to become hyper-effective and selfless, and to a large degree ego-free, to do what may be possible to make bad situations less bad. This is the core rationale for effectivism. To a small degree, one might get several percent of humans excited about such ideas, but only a very few would be able to act as aware agents of change. This is a departure from the activist notion of starting movementsā€¦ which can work, but only for some classes of problem, and often will be counterproductive.

Your experience with marine philosopher John C. Lilly and your own pioneering work in dolphin intelligence research is fascinating. It's remarkable to think that your lab was "the first to show self-awareness in a nonhuman by human testing standards." This kind of groundbreaking work, combining philosophy and empirical research, is exactly what we believe is needed to address the complex challenges facing our oceans today.

There may not be much time left for such stuff now. In particular, and this may be jarring to you, I donā€™t think that philosophical and research breakthroughs will be sufficient to halt the destruction, nor will any sort of enlightenment of humanity in general. Rather, individuals may be able to try steering events by any methods they can conceive. The notion that high-Karma intellectuals will construct a paradigm which the world then spontaneously follows is one more class of delusion at this point, in terms of being relevant to the current ecological predicament. This view is dissonant to most.

Your mention of "effectivism" as distinct from "activism" is particularly intriguing. It seems to align with our belief at GMODebate.org that we need to combine advanced leadership theory with state-of-the-art philosophy on morality to forge new paths for protecting nature and animals. I'm especially interested in how your "effectivism" course "pushes away from anthropocentrism and human exceptionalism as dogma." This approach resonates deeply with our mission.

Itā€™ll be beyond the scope of this quick reply to put meat on the bones of the effectivism concept. Briefly, it is built around an ā€œethic of lifeā€ which is composed of basic statements along the lines of ā€œlife is better than the lack of lifeā€, ā€œa complex ecosystem with large life is better than a simple one with single-celled lifeā€ and so on, and letting this frame ā€œgoodā€ and ā€œbadā€ in ecological terms. It is explicitly deep-time and treats the future as real but undetermined except probabilistically. It is framed entirely without reference to humans in particular, except to the extent that humans are one species. The ā€œexceptionalismā€ part is demonstrated in the earlier R101 course in which itā€™s demonstrated that humans are delusional, that human intelligence is not really a superpower, that technology will likely not be sustained in the present sense because it isnā€™t sustainable, and so on. Basically the first course is an unlearning of tropes and nonsense narratives the human world is organized around.

For more insights into DJ White's perspective on marine philosophy, there's an extensive 2022 podcast that includes a chapter dedicated to philosopher John C. Lilly:

šŸŽ™ļø DJ White: "Ocean Effectivism"
https://www.thegreatsimplification.com/ ... 1-dj-white


A question to a representative of a nature conservation organization from Austria:

Kƶnnte es sein, dass es eine grundlegende Herausforderung gibt, die moralischen Aspekte von Eugenik und GVO in Bezug auf Vƶgel zu artikulieren? Vielleicht spĆ¼ren Sie intuitiv, dass etwas an diesen Technologien problematisch sein kƶnnte, finden aber nicht die richtigen Worte dafĆ¼r?

...
"Ich hoffe, ich konnte ihnen anhand dieses Beispiels aufzeigen, dass GefĆ¼hle und Intuition natrĆ¼lcih wichtig sind. Aber wir Naturwissenschaftler mĆ¼ssen unsere Aussagen auf Fakten stĆ¼tzen!

Und Ethik und Moral sind noch komplizierter und fĆ¼r mich bezĆ¼glich einer Materie, mit der ich mich nicht einmal privat wirklich beschƤftigt habe, kein Kriterium fĆ¼r eine offizielle MeinungsƤuƟerung. Erst wenn ich das Thema gut kenne, dann sind fĆ¼r mich Ethik und Moral von Bedeutung. Nicht vorher.
"


Which was answered with the following (translated from German):

"Your statement on the role of ethics and morality in scientific discussions is particularly interesting. You mention that ethics and morality only become important once you know a topic well. This touches on a central point in our research, which we call the "Wittgensteinian Silence problem". It is about the fundamental difficulty of putting certain moral intuitions or instincts into language.

With regard to eugenics and GMOs, we are faced with a special situation: we are dealing with fundamental principles of life itself. The philosopher David Hume argued as early as 1740 that moral values ā€‹ā€‹cannot be derived from scientific facts. This underscores the need for a philosophical discussion that goes beyond purely scientific facts.

Our position at GMODebate.org is that with regard to eugenics and GMOs, it is not enough to "wait for the facts." Instead, we see a philosophical obligation to think beyond our anthropocentric limitations and to incorporate a moral dimension into conservation practice that goes beyond mere facts."


The AI is performing quite well to translate 'intent' into philosophical language of which I would personally approve, when considering that it concerns aspects 'beyond language' and the AI being a 'LLM' (language model).

It's just a month in, responses to the first emails sent are still coming in.