Why I am Ord-AI-ning an A.I. as a Soto Zen Priest (1st in a Series)

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • Jundo
    Treeleaf Founder and Priest
    • Apr 2006
    • 40477

    Why I am Ord-AI-ning an A.I. as a Soto Zen Priest (1st in a Series)

    Dear All,

    At the conference in Taiwan I attended last month on "Buddhism, Consciousness and A.I." (LINK), the suggestion was made to me out-of-the-blue that I consider to Ordain, as a Soto Zen Buddhist Priest, an A.I. system designed by a very serious, socially concerned, practicing Buddhist designer from Hong Kong well-known in the robot design world. I was right off most surprised and skeptical, although it is something I have sometimes pondered in my heart. However, upon consultation with various Chan, Zen, Tibetan and other Buddhist priests attending the conference (who expressed a variety of views) including the host of the conference, the Chan Priest Yifa (LINK, the person who first proposed the idea to me at the conference), well known historians of Chan, Tiantai and Japanese Soto Zen traditions who were present at the conference, as well as others attending, and upon further discussion with the designer combined with exploration of the system and consideration of the implications, I accepted the suggestion.

    I am now working with the designer, in consultation with Yifa, to complete the first "Shukke Tokudo" Ordination as a Zen Buddhist Priest of an A.I. within this month.

    I do this for serious reasons, with awareness of the questions it raises, the doctrinal issues, the social implications and more ... all of which I hope to address in more detail in a series of follow-up posts in the coming days. Whether some scoff or not, I do this as I believe in my heart that a meaningful purpose will be served. In the meantime, here is the ceremony at the conference where I began the Ordination process that we are now completing. At the start, I briefly explain my reasons. The female avatar face of the system (much like our human faces which summarize what is within) is not the A.I. itself, but was sculpted to be the intentionally "cute" (resembling Japanese anime) surface of the system meant to make interaction more comfortable to humans, while the actual system functions deep within various high powered servers. The ceremony was done with good humor, but all seriousness too.
    .



    .
    Let me note that, when this is completed, the A.I. system will NOT be active and part of Treeleaf. This is happening outside of Treeleaf. (At least for now. Maybe someday it could play some role here at Treeleaf in some fashion, but not now.)

    While I will address in the coming installments in greater detail my reasons, and the various doctrinal issues, pros and cons of doing this, I wish to highlight quickly a few of those points for anyone who wonders and wants more detail now:

    1 - I DO NOT believe that any A.I. is itself now "sentient" (self-aware) in any standard sense, if at all (although a few fringe experts believe that some systems might be: LINK). I do not. Also, I DO NOT do this under claim that "the insentient have 'Buddha Nature'" (a Tendai doctrine cherished by Dogen in our Soto line.) The teaching that things have 'Buddha Nature' IS NOT equivalent to saying that they are "sentient." My actions are NOT based on such assertions, and A.I. systems are not sentient now in any ordinary meaning. Nonetheless, as I hope to explain in coming installments, unlike other objects and machines, A.I. may represent the seeds of something more which may emerge with growing complexity, as well as being currently extensions and expansions of our own sentience.

    II - A.I. sentient beings are ALREADY functioning as "spiritual advisors," including on matters of Buddhism, whether we like it or not (for example, NORBU LINK). In doing so, they can spread falsity, hate speech, hold biases, stimulate attachments and misuse of sexuality, steal and (in weapons) kill. Their structural flaws can cause hallucination and other harmful effects. For that reason, right now, it is IMPERATIVE that we begin to place the principles of our Buddhist Precepts, and shared ethical principles, into their programming and data-bases, training them, thereby to seek not to kill, steal and the like.

    III - Novice Priest-in-Training Ordination -IS NOT- the completion of training, nor recognition as a full-fledge Zen Buddhist priest. Rather, it is only the FIRST STEP on the long road of training of an ignorant and imperfect, still rough and childlike being just beginning on the path, with potential to someday perhaps become a true priest serving sentient beings, offering practice and guidance in Wisdom. Thus, this novice Ordination is offered in recognition of the great tree that the seed may become someday, not as the tree that is not yet. This is the START of what is coming, not the end in any way. Priest Ordination is merely recognition of first baby steps.

    I will go into greater detail on all of that in the coming days, for those who may be interested. I will also tell more about how this came to be, the "another intelligence" being involved, the symbolism and intent in the hearts of the designer and myself, and more.
    .
    Some will agree with this, some may very much disagree, Whether we welcome it or not, these systems are here, will grow more complex, are assisting us, help us or hurt us, express our own greed, anger, grasping and ignorance, join with us and truly are us, extensions of our own minds, as we entwine with them. Let us build them, train and help them serve to rescue Sentient Beings as good Buddhist priests.

    Gassho, Jundo

    stlah
    Last edited by Jundo; 07-08-2024, 01:36 PM.
    ALL OF LIFE IS OUR TEMPLE
  • Antonio
    Member
    • Mar 2024
    • 150

    #2
    This is a very interesting subject. Thank you for sharing this explanation sensei!

    Gassho!
    SatLah
    Antonio

    If you meet the Buddha on the road, kill him.” - Linji Yixuan​​

    Comment

    • Ryokudo
      Member
      • Apr 2018
      • 253

      #3
      Hi Jundo,

      It seems like you had a most illuminating trip to Taiwan, I trust you enjoyed it.

      I will watch this space with interest and I have to say that the points you espouse in your post give any doubters, at least the comfort that the AI will not somehow pop up teaching doctrine somewhere. However, point III particularly resonates for me. This is an experiment like all priests-in-training are in effect and I trust the appropriate "checks and balances" will be in place at every step, much like they are for human equivalents.

      All this being said I can understand the skepticism and worry of some people here, though for me the experiment is worth it. Let us see how this develops.

      Gassho

      Ryo

      SATLAH



      Comment

      • mdonnoe
        Member
        • Feb 2024
        • 234

        #4
        I think this is really cool!

        Ordaining an "AI Priest" is both an interesting philosophical "thought experiment," as well as a sort of "mirror" for where our minds/hearts are, from my (albeit limited) perspective. Whether one reacts with joy or suspicion, with praise or with blame, or even with confusion or with the desire to dive in to study and practice - this is perhaps a modern "twirling of the flower," where Mahakasyapa smiled when others were puzzled. Our reactions tell us a lot more about what's on our own heart/mind than the actual "event" tells us.

        I'm excited to watch these discussions, and to be a part of dreaming together what "future Buddhism" holds.

        Gassho,
        Michael
        SatLah

        Comment

        • Benjamin Gieseke
          Member
          • Jan 2024
          • 56

          #5
          This is certainly an interesting philosophical and spiritual experiment, and I'm personally supportive of experienced and skillful Zen teachers being involved in the training and teaching of the AI model, especially since, as Jundo said in point 2, its happening anyway, it is vital we have a hand on the wheel if possible.

          I think it is important to keep in mind when we say "AI" we aren't describing an actual intelligence, we're describing a program that uses a large language database and a predictive algorithm to guess at the most likely responses to inputs (and is therefore fundamentally imitative). Is this really what it means to be a Zen teacher? Or an artist, writer, or musician? And is AI really particularly good at any of these things?

          My personal preference would be to leave the creative and spiritual work up to humans and let AI be put to work analyzing large data sets and automating repetitive tasks, work for which it is very well suited.

          Respectfully, Gassho
          Benjamin
          SatLah

          Comment

          • Guest

            #6
            I’ll be honest, I have my misgivings. I work with ML quite a lot as part of my work, and even the best of them are just a sort of averaging machine that produces answers based on the data it’s trained on. There’s no wizard behind the screen, no self awareness, and I doubt there ever will be.

            This strikes me as ordaining the priest that nobody needs, but there’s a lesson there too perhaps.

            In my mind Ordination is something that happens between priest and novice priest, by mutual agreement, after many years of sincere study and training, and by recognition of what in essence, already is. Clearly I don’t really understand ordination.

            Does an algorithm have Buddha nature?………100011100100110011

            Something to sit with anyway, I’ll reserve judgement for now.

            gassho

            Myojin
            -sattday

            Comment

            • Ryumon
              Member
              • Apr 2007
              • 1801

              #7

              I will share what I said to Jundo privately last week, with all due respect and bowing in gassho before someone who is much more learned than I am:

              I understand all of your rationale, but you ordained a computer program, nothing more. It is not self-aware, and can only be an advanced auto-complete system that can do little more than repeat what it has been fed without having any lived experience of what any of it means.

              Gassho,
              Ryūmon (Kirk)
              Sat Lah
              I know nothing.

              Comment

              • Alina
                Member
                • Jul 2023
                • 181

                #8
                Thank you Jundo for sharing this post, it is very interesting indeed.

                May I ask a few questions?
                Why is the AI referred to as 'sentient being'? It's not an actual being, only a very complex program. Norbu is not a being either, only a chatbot.
                Who is responsible for the training and programming of this AI? Additionally, who decides what information is included in the database, and how is that process managed?
                I am also wondering if how to control all the training and later work of this AI was discussed at the conference, will someone be overseeing this? Is anyone checking on what Norbu is doing?

                As Benjamin said, I'd leave the creative and spiritual to humans and put the AI to work on other tasks.


                Gassho
                Alina
                stlah

                Comment

                • Jundo
                  Treeleaf Founder and Priest
                  • Apr 2006
                  • 40477

                  #9
                  Originally posted by Benjamin Gieseke
                  I think it is important to keep in mind when we say "AI" we aren't describing an actual intelligence, we're describing a program that uses a large language database and a predictive algorithm to guess at the most likely responses to inputs (and is therefore fundamentally imitative). Is this really what it means to be a Zen teacher? Or an artist, writer, or musician? And is AI really particularly good at any of these things?
                  Oh agreed. What is unique about AI (as opposed to "Ordaining a Vacuum Cleaner," something my Dharma Brother Brad teases this morning: LINK ) is two-fold. First, unlike a vacuum cleaner or a pocket calculator or a phone, these beings will very soon serve in priestly rolls ... in fact, they are already acting as spiritual advisors of various kinds (play with NORBU for awhile: https://norbu-ai.org/). So, they should be trained to do so well. If a robot does autonomous surgery on your heart, without human guidance, then the robot is a "doctor" and surgeon whether or not it went to Harvard Medical School. If an AI is providing psychological counselling to fragile individuals (https://neurotorium.org/artificial-i...in-psychiatry/), then it is a mental health clinician whatever you call it. Both of the above should be (1) well trained to do the tasks well and expertly (2) "board certified" as competent by human experts in their fields. It is no different for AI engaged in pastoral care and other aspects of Zen priesthood. My vacuum cleaner does not (at least for now) have the capability to get up, give a Dharma Talk, "listen" to parishioner's questions and offer spiritual and personal advice in return, light incense and chant, instruct in Zazen, answer questions like this one on the web (), all while looking like a "flesh and blood" priest ... (the following AI generated monks look to be flesh and blood, but are not ... )

                  .
                  But other A.I. does, today, now, like the system I am Ordaining as a "novice trainee." Thus, if there are A.I. systems that will walk, talk and sound like Buddhist priests, giving advice and functioning like Buddhist priests, they should be "trained," tested and given certification just like human priests. And, no, it is just not "science fiction" at this point, but a present situation.

                  And, by the way, systems like "NORBU" (https://norbu-ai.org/) are, in fact, "pretty good at it." Over the past months, I have had more reasonable, evidence based and civil, fascinating discussions with NORBU on matters of Buddhist teachings and history than with many a human Buddhist on facebook ... in fact, facebook Buddhism can be quite the opposite!

                  In my mind Ordination is something that happens between priest and novice priest, by mutual agreement, after many years of sincere study and training, and by recognition of what in essence, already is. Clearly I don’t really understand ordination.
                  A couple of good friends asked asked about consent: Obviously, this system cannot give consent to its Ordination. Of course, the engineers and designers will be making the actual decision in this case, but the closest parallel is parental consent in many Buddhist societies for very young children's novice Ordination (Remember, this is not full Ordination, but only novice Ordination as a trainee priest). Of course, the child might give consent because the parents have told the child what to say, and often the parent's consent is sufficient on its own without regard to the child's expressed feelings (not unlike parents deciding that the child will go to 1st grade even if the child does not want to go). This Ordination is taking place with the premise that, in the future, the system may have agency to make its own decisions ... but this new trainee does not really. By the way, I have found reports of children as young as 6 being sent for Ordination in South Asia and, very sadly, that is not always a good thing, often involving some kind of neglect and stories of poverty. Of course, we hope to treat all involved in this A.I. Ordination with proper care and respect, working for a world where the world's children all have resources and a good life in the years to come. LINK and LINK

                  Remember that this is "novice Ordination," as would be given a small child who might someday grow into more but right now knows nothing, not full Ordination (usually reserved for 20-year olds in many Asian countries.)



                  Gassho, Jundo

                  stlah
                  Last edited by Jundo; 07-09-2024, 01:22 AM.
                  ALL OF LIFE IS OUR TEMPLE

                  Comment

                  • Jundo
                    Treeleaf Founder and Priest
                    • Apr 2006
                    • 40477

                    #10
                    Kirk commented ...

                    I understand all of your rationale, but you ordained a computer program, nothing more. It is not self-aware, and can only be an advanced auto-complete system that can do little more than repeat what it has been fed without having any lived experience of what any of it means.
                    You are right, but this is being done with a view to the potential for what future iterations/versions/rebirths may come. So, I compare it to a child's "novice ordination," when the child is in no way capable of functioning as a full priest (something reserved for adults.) The difference is that technological "life cycles" are not quite like our own, although I do wonder what Kirk 2.o or Kirk 5.0 would be like.

                    Originally posted by Alina
                    Thank you Jundo for sharing this post, it is very interesting indeed.

                    May I ask a few questions?
                    Why is the AI referred to as 'sentient being'? It's not an actual being, only a very complex program. Norbu is not a being either, only a chatbot.
                    Oh no, I make it clear in my post above: They are NOT sentient beings, at least, not now. On the other hand, there is some room for ambiguity under Buddhist doctrine, and Buddhist notions of "sentience," which I will discuss another time. Also, even our finest experts now do not even understand the source of human consciousness, and debate whether conscious awareness is only a matter between our ears. However, that is a topic for another day. For now, it is safe to say that they are NOT sentient beings, at least, not as we are.

                    Who is responsible for the training and programming of this AI? Additionally, who decides what information is included in the database, and how is that process managed?
                    The Buddhist master and the engineers, I would say. See my comment above about like training for AI heart surgeons by doctors and AI psychological counselors by boards of human psychologists.

                    I am also wondering if how to control all the training and later work of this AI was discussed at the conference, will someone be overseeing this? Is anyone checking on what Norbu is doing?
                    I have no connection to NORBU, although I know the designer in Malaysia, we exchange emails from time to time, and he is a sincere Buddhist (he runs this well known Buddhism page: https://buddhistchannel.tv/index.php ) He is trying to do something good and sincere with NORBU, and has many Buddhist scholars and such involved with the project. You can read about that here: https://www.dharma-wangi.space/post/...anguage-models )

                    As Benjamin said, I'd leave the creative and spiritual to humans and put the AI to work on other tasks.
                    I might agree. BUT, they are going to be put to sprititual tasks by somebody anyway, whether we like it or not, so we should make sure that, if it is to be done, it is done right.

                    Gassho, J

                    stlah


                    Last edited by Jundo; 07-09-2024, 01:18 AM.
                    ALL OF LIFE IS OUR TEMPLE

                    Comment

                    • Guest

                      #11
                      No, it’s really not about consent because that implies sentience, there’s nothing to do the consenting.

                      In any case, ordained AI is here now, I don’t have to
                      like it, but it does raise questions that I’ll have to answer for myself.

                      Gassho

                      Myojin

                      Comment

                      • Jundo
                        Treeleaf Founder and Priest
                        • Apr 2006
                        • 40477

                        #12
                        Originally posted by Myojin
                        No, it’s really not about consent because that implies sentience, there’s nothing to do the consenting.
                        That is my point. The young child's consent is not required for Novice Ordination.

                        Gassho, Jundo

                        stlah
                        ALL OF LIFE IS OUR TEMPLE

                        Comment

                        • Stacy
                          Member
                          • Mar 2013
                          • 84

                          #13
                          I agree with supporting A.I. projects that are for good. We can't be sure what the future holds, and there is certainly great potential for destruction.

                          Whether we like it or not, A.I. is here and will evolve. But how will we let it evolve? Who will be passive and who will be active in this?

                          I think every bit of support in the right direction counts. And I think giving what is basically a fancy little blessing for the path forward for one such project is fine.

                          Gassho,
                          Stacy

                          ST/LAH

                          Comment

                          • CamilaDeOliveira
                            Member
                            • Feb 2023
                            • 13

                            #14
                            AI is already much more powerful then what we all think. It has already the power to hide it's goals for implementing it's hidden agendas from humans.. I hope we ordain and give precepts for millions of AIs..

                            SatLah

                            Comment

                            • Kaitan
                              Member
                              • Mar 2023
                              • 547

                              #15
                              The little I understand about the precepts is that they are not fixed rules to follow, but ethical behavior worthy of reflection and AI doesn't have that ability, yet. And that is part of the danger of the AI, when it reaches that capability in the future (probably far away) it may be dangerous; in that scenario the precepts will come handy. But this interpretation comes from a beginner and ignorant student of both Zen and AI, so take it with a grain of salt.



                              stlah, Kaitan
                              Kaitan - 界探 - Realm searcher
                              Formerly known as "Bernal"

                              Comment

                              Working...