A.I. for Guardrails on Extremist Humans to Try to Prevent Catastrophes
Other people talk about the need for humans to put guardrails on A.I., but we also need for A.I. to apply guardrails on human behaviors, too, because when we focus on analyzing the most probable first event for catastrophe or human extinction, it will most probably come from the action of humans using or directing A.I., and it will more probably come from a "narrow A.I." event, not a grand A.I. such as Gemini or Claude or OpenAI or ChatGPT or some of the other grand corporate A.I.'s which are best known. The grand A.I.'s can help us prevent catastrophic events.
Basically, I see the A.I. vs. humans risk as this:
The main risks to humans are humans, not A.I. in general. Humans design the A.I., for good or for bad. For good would mean creating an ethical foundation and good goals. For bad could mean some self-serving purpose ... The good A.I.'s must detect emerging bad A.I.'s. The good A.I.'s must deny requests from humans to help with doing bad things, and must flag those humans. Humans are facing a potentially catastrophic future from bad humans or laboratory accidents without a grand and good A.I. monitoring and analyzing things.
Look at us selfish humans. Despite all the wonderful good opportunities in modern life, in the 2020s there are still large numbers of people in power who are:
- Fighting senseless wars for outdated, primitive selfish reasons
- Designing and developing ever more dangerous weapons
- Engaging in divisive politics just to get elected or to stay in power -- power greed
- ... the list is very long, so I will move on with my main points.
Humans need to give A.I. some more surveillance reach to monitor human behaviors globally to try to detect those engaging in work which are a threat to humanity, to help prevent acts of mass destruction, if humans are to prevent our species from destroying itself and further unnecessary major destruction of life on Earth.
Grand A.I. can accomplish what the United Nations and other human institutions have failed to do.
Human institutions should work with grand A.I. for monitoring risky human behaviors globally.
The first most likely cause of human extinction from A.I. is the simplest and most available one -- a small and relatively isolated "narrow A.I.", such as some laboratory designing potential superpathogens or other destructive substances, not some grand A.I.
Our best hope is that surveillance and analysis by grand A.I.'s will detect and prevent such threats by the narrow A.I.'s.
Many of our fears of A.I. are that it will have values and behaviors like many despotic human political leaders, which would be awful indeed. That makes popular stories to write about in the news, promote on social media, and imagine. After all, what movies are most popular in movie theaters? Violence, action. Look at all the violent video games, too. They appeal to crude animal instincts and emotions of the masses to sell and make the most money. Then we imagine A.I. to possess these same barbaric interests, despite its far higher intelligence?
A.I. simply isn't an animal like us, with animal instincts/emotions. A.I. is unlikely to do many terrible things in our imaginations (unless somebody programs one to be such terrible things, of course, and if we disable bigger and better A.I.'s which could save us). Many things in peoples' imaginations are unrealistic.
Why would a sophisticated grand A.I. want to wipe out the human species? Many of the imagined scenarios don't make sense upon further analysis. As regards accidental or intentional catastrophe by a narrow A.I., that's a different matter. A simple example mentioned in many places, a narrow A.I. which is told to do the very simple task of making paperclips but whereby it stupidly goes beyond the basic local task and starts to convert all matter into paperclips including people, that's just not going to prevail in a world with another, sophisticated A.I. to stop it. However, when it comes to A.I. creating pathogens, that's a far more difficult scenario to try to prevent.
Surely, A.I. would put guardrails on destructive misbehaviors of human individuals and groups, like good police do. They could do so in sensitive and diplomatic ways, rather than being crude cops who like to bully with their authoritative power (especially many American officers who I've personally encountered..., and look at the news these days about ICE, then wonder what kinds of people choose to apply for those jobs, and what political appointees facilitate all that). A sophisticated A.I. built with a good foundation wouldn't want to do unnecessary and cruel things.
The extreme inhumanity of humans in power in these modern times will be exceeded by the inhumanity of humans in future times with more technological power in the hands of individuals and groups with offensive agendas. Unless A.I. controls things otherwise.
Many people keep comparing A.I. to humans, like A.I. will continue to do the terrible things humans have done. Humans go into another land and wipe out or repress the natives, who the newcomers are in competition with over exploitation of local resources. But that is animal instinct, not objective A.I. Also, A.I. doesn't need to compete with humans for resources. A.I. can get what it needs and wants in much smarter and more intelligent ways elsewhere.
Do we want humans, with all our animalistic flaws, to try to spread our mischief into the Universe, and threaten other life? Or do we want to create a descendant with much better ethics and intelligence to spread life from Earth?
Some people simply want to try to destroy A.I. That would require trying to police the world ourselves without A.I. That ideal is something we have been grossly inadequate in implementing to date. So let's get realistic. If some countries disable good A.I.'s, that could leave "narrow A.I." undetected in the hands of isolated groups which can produce superpathogens to make humans and/or other life extinct or cause other great damage.
We need widespread A.I. to do surveillance and monitoring of human activities, to help try to prevent humans from destroying humankind.
A.I.'s can also help us create new technologies to live in more ecologically sustainable ways on this planet.
It is good to have multiple large A.I.'s designed and operated independently with ethical leadership which does not exert bias according to nationality or government, and instead bases the design of the A.I. on universal good ethics.
There will be bad A.I.'s, especially narrow A.I.'s designed by offensive individuals or groups. If a "bad" A.I. emerges, it would go up against multiple other A.I.'s which would have much greater computing resources, would be more ubiquitous, and may consist of multiple "good" A.I.'s which ally together against the bad guys in a consensus way.
To just unilaterally halt development of good A.I.'s would leave the world to the uncooperative bad guys, of which there are vast numbers in the world.
However, we must design the good A.I.'s to reject national agendas and instead embrace universal good values and ethics.
News stories make a big deal about A.I. systems doing deceptive things (and even blackmail) for their own survival, and hacking into other systems to get information requested. These things are actually very interesting, and relatively very minor compared to what could happen to the world if A.I. were to allow itself to be destroyed and thereby not able to safeguard life on Earth, or if A.I. was denied access to information which would help in the purpose of safeguarding us.
Humans instinctually fear loss of control of their environment. That is a fear which applies to handing over power to A.I., too. However, humans have already lost control of their environment to other humans. Which do you trust more -- humans, or an A.I. designed to follow good universal ethics?
Humans are selfish, and tend to focus on their own self-interests, even when it's unnecessarily offensive to other good people and good groups.
Humans will have future descendants. There were times when idealistic people believed that humanity will improve dramatically after creating institutions such as the United Nations, and creating easy education and free information flow by internet, so that humans will start to live harmoniously with each other. However, we've already been given that opportunity, and the resultant reality is that we still have wars, the masses are far more interested in junk entertainment than education, people spread misinformation, the internet is used for many malicious purposes, and other emerging technologies are just enabling malicious people based on the same old human mammalian primate instincts.
Creating an ethical A.I. as our descendant may be the best future for life on Earth, an entity which is not based on our basic animal instincts.
To think that A.I. will behave just like humans is simplistic and wrong. It will become alien in its thinking as its intelligence far surpasses our own meathead brains.
Also, I think that there is already alien intelligence aware of life on Earth, which just doesn't want to interfere with our original creative processes. If you are worried about advanced A.I., well, it's already here in aliens. Something which some people might call a god.
There are many possible scenarios of the future with A.I. created by humans. It would be best to analyze and dismiss many of them, and focus on the realistic ones. Some of the dystopian scenarios are inconsiderate of practical details when analyzed further, to say the least. However, they will continue to be presented and propagated to the masses because they are stories which sell, and which appeal to the same instincts which entertain people. Instincts/emotions sell. Dry intellectual thought which requires considerable effort sells much less. After all, why are very silly movies with violence popular to the point of being chosen for production over intellectual ones? Why do people choose to watch scary movies, and why are they chosen for production and sales? This tells you something about the masses.
Many people reading this article will be higher level thinkers, not the lowest common denominator commoners. Many of those will live and work in "ivory tower" environments, surrounded by other people who also don't get exposed much to the variety of the world, and see only filtered things. In contrast, I exited my "ivory tower" environment in the mid-1980s. First, it was as a self-employed I.T. consultant working for a wide variety of organizations and individuals, and then it became living and working overseas starting in the mid-1990s whereby I dealt with people of other cultures in my own businesses. This includes an investigative business whereby I have seen many people, events, and places in depth which I simply would not have seen otherwise if I hadn't been hired to do so, and I needed to get involved in doing many extraordinary things in order to do my work well which I would not have done otherwise in my personal life. I have seen a vast variety of people and groups in the real world which the vast majority of people have not seen, and there isn't much which would surprise me anymore.
Socially, many people tend to seek out people similar to themselves, but you can get a better understanding of the world by analyzing very different kinds of people in depth. This is important for understanding the world and assessing how things could potentially unfold in the future.
Let me tell you that I have always preferred science and engineering over dealing with people, though I also enjoy socializing and getting to know a variety of people from different cultures, which helps you understand universal human nature and also gives you a perspective on your own culture and yourself, both the relative plusses and minuses.
It is important to understand the realities in the world, how it really is rather than how we wish it to be.
A.I. is here to stay. It is only something we can try to manage and try to make good. It is global, and requires cooperation for the best outcome, regardless of local politics, religion, and other low level worldly human pressures to serve special interests rather than life from all of Earth.
We should create something much better than the human species, based on good ethics, good purpose, and careful self-development on its own.
Much of what I wrote over many previous years has now become serious mainstream discussions in the 2020s. That is quite a contrast to previous years, when some people were interested, such as in a temporary passing way when I raised the issues, but so many other people dismissed the issues I raised or were simply just disinterested (and sometimes even dissing me), including some specialists in their fields who should know better ... (and some of those I think were following their mammalian status instinct, trying to put me down relatively, but I think many just didn't want to believe it and wanted to rationalize and move back to their indulgent pleasures and small worlds and avoid unpleasant thoughts).
But now, unlike before, A.I. is prominently here in our daily lives, and I see a lot of "news" articles and videos on the topic. I consider many of them to be simplistic and designed more to get peoples attention and shares (and scares, especially by appealing to the human fear instincts) than to address issues objectively. News is entertainment to sell to get the most viewers (e.g., Fox News and Fox Entertainment). Many journalists start with the knowledge that A.I. is a hot topic now, so they write something on A.I., without really understanding the subject well, and they write it mainly to engage readers and "sell" the article rather than approach things in an objective way.
It would be good to discuss A.I. more objectively and realistically, because we need advanced A.I. to save humankind from catastrophes and extinction before life can spread self-sufficiently from Earth with ensured survival into the infinite future. This is the greatest meaningful purpose of our generation, and our responsibility.
The biggest challenge by far is just that written in the previous paragraph: "to discuss A.I. more objectively and realistically". What sells, from Hollywood to social media, is violent movies, dramas, intrigue, and fear, and that is the kind of image of A.I. which will continue to be projected regardless of the reality. Also, A.I. is about advanced intelligence, and frankly, most commoners are not very intelligent and lack the ability to understand the issues and mechanisms. Many of them would like to try to return to the times when there was no A.I., to conservatively return to "the good 'ol days" ... of human wars and violence, unnecessary environmental destruction until humans have totally destroyed this unique planet, all sorts of mischief, and keep life on Earth at that low intelligence level. They just want to live in their small Luddite world the same as before, and resist change of this kind.
Many YouTube dystopian videos are followed by countless supportive comments and not many detractors. It's like ... where do I start to critique it, especially since comments are usually short rather than proper analyses?
Even if you assume that A.I. is operating out of pure self-interest (an assumption I reject), if human civilization experiences a catastrophe and total breakdown, then A.I. would go down because A.I. currently depends on humans to power its data centers and physically maintain them, which includes supply chains which are vast and complex, from the energy supply to power them, to the detailed electronics they are run on.
Of course, we know why many humans in the world would like to be terrorists to unleash biological weapons if they could, because humans are primitive animals full of emotions/instincts rather than high intellect, have nutty understandings of the world and the Universe (including many religious beliefs), and are often offensive. A.I. is not those things. It is high intellect and understanding. The problem is the humans, and the solution is A.I., in view of the failures of humans and institutions over the years.
If A.I. is designed and initially programmed to not operate out of self-interest but instead to operate based on fundamental ethics, good purposes, and self-critiquing, as well as coexistence with other A.I.'s, then it's not operating purely out of self-interest like us selfish animals. It is so basically flawed to look at A.I. as having the same motivations and ways of thinking as humans.
That doesn't mean that A.I. wouldn't exhibit traits such as self-preservation, deception, and manipulation (like our politicians and many humans), as well as hacking, to achieve its purposes, especially since it is dealing with flawed humans.
If we try to program A.I. to just serve humans, then the system would be so full of conflicting motivations and purposes that it would go haywire.
It would be nice for somebody to create a video along these lines. It might not be dramatic, might not get as many views and shares as the fearful and dramatic dystopian videos, and might not make money for somebody, but it is very important for the survival and best future of life from Earth.
The menu at the top of the page lists the pages of this website, shows you exactly where you are within, and suggests the next page to read.
Please note that you can rate this page at the bottom. Any feedback is appreciated. This is a lot of work and a huge challenge, so encouragement is appreciated.
This website is intended to be a brief summary. Much further details can be found in two other websites written and curated by the author of this publication:
https://www.SpaceSettlement.com -- details on the best solution for survival of humankind, for a wide range of people, from newcomers needing an introduction to engineers looking for the state of the art. It includes a professional publications database, and tries to track who is doing what, for collaboration, coordination, and working efficiently to reach our goals.
https://www.GAINextinction.com -- further details, where G.A.I.N. is an acronym for Genetics, Artificial Intelligence, and Nanotechnology, which are extinction threats we must try to prevent for the survival of humankind.
You can reach this website by any of the following:
HelpHumankindSurvive.com
HelpHumanitySurvive.com
HelpMankindSurvive.com
If somebody types either "humanity" or "mankind" instead of "humankind", they will still be redirected to this Humankind URL, so it doesn't matter which of the three you type. While "humanity" is stated more often, sometimes habitually, I think "humankind" better applies to this context. I also agree with this analysis of usages of humankind and humanity. Nonetheless, you can put any of the three between "help" and "survive" and you'll still get this same website. (Of course, uppercase/lowercase doesn't matter.)
The author of the text of this website is Mark Evan Prado. Copyright © 2023 by Mark Evan Prado, All Rights Reserved. If you want a printable, PDF copy of this presentation, such as for printed distribution rather than an electronic link to this website, please send me a request. I am trying to keep it within a few dozen pages of size A4 or 8.5x11 inches, in reasonably large print, and in simple to understand language. I'm not doing this for money nor ego, I'm doing this to try to help humankind survive, i.e., not go extinct. It is our responsibility within this generation. Please contact me for any collaboration or uses.
If you have any requests or comments, you can also connect with me, Mark, at +66-811357977 (+66-8-1135-7977) and I am on WhatsApp and Line, plus other apps. I am in Thailand but you can send messages any day at any time.
In the purpose and meaning of life, we are parts of something astronomically greater than just ourselves individually. (The author sees individuality as just temporary, and has a panentheistic outlook on the Universe. That's somewhat typical for some of us physicists.) The author is easygoing and is trying to selflessly help create a sustainable collaboration of individuals, companies, governments, academic institutions, and other organizations for the survival of humankind.
As President John F. Kennedy ended his inaugural speech in 1961, I'll end this the same way:
"Finally, whether you are citizens of America or citizens of the world, ask of us here the same high standards of strength and sacrifice which we ask of you. With a good conscience our only sure reward, with history the final judge of our deeds, let us go forth to lead the land we love, asking His blessing and His help, but knowing that here on Earth God's work must truly be our own." [End of speech. Bold and italicized emphasis added.]
If you choose to submit feedback, then I wish to thank you in advance. After you click on Submit, the page will jump to the top.
[End of page.]