Roshi continued his discussion this week of Artificial Intelligence (AI) and how it relates to our practice.  There has been a lot of talk these past few weeks about AI language models rivaling human abilities.  Some examples include writing essays which would be accepted in an undergraduate college course, generating complex computer code from English requirements, and even writing poetry.  To test the latter, the sangha asked an Internet based AI to write a Haiku of thanks for Roshi.  This following was presented after the talk and Roshi was actually impressed by the quality of the poem.


Gentle guiding hand 

Leading us to new horizons 

Thank you, dear Roshi


Goroshi asked the question – what does AI have to do with our practice and could a computer ever realize Buddha nature?  To really consider this, we need to think about the nature of data, knowledge, and wisdom.  If we think along a continuum, on one side we have data and on the other extreme we have wisdom.  Between these two we have knowledge.  Data is simply individual facts stored in isolation.  “Albany is the capital of New York.”  Our initial response to a fact is always, “so what?”  This is because there is no context to this fact.  Humans take data and use it to generate new knowledge.  For instance, knowing that Albany is the capital of New York, we can infer that there are many politicians in Albany.  We can next infer that if we want to lobby for a cause, we may wish to travel to Albany instead of some other city in New York state.   In essence, humans can take data and apply it to solve problems.  This turns data into usable knowledge.  


When we consider modern computers and the Internet, we understand that AI has access to much more data than any single human.  In fact, we can assume that the largest AI models have access to the entire corpus of the human written word.  Networked computers also have access to enormous amounts of computational power.  It is this power which is now allowing AI’s to begin applying data to potentially generate knowledge.  Roshi pondered whether the emergence of real knowledge from AI computer systems would apply economic pressure to “knowledge workers” in a similar way to how robotics applied pressure to factory workers.  He also wondered if this is why there is so much uneasiness about this new technology.


However, our issue is not economic, it is about the other end of our continuum – namely wisdom.  Could an AI computer system leap from generating knowledge into having true wisdom?  Roshi started out by stating that this is a very hard question.  For instance, at its heart, AI is a simulation technology.  This means that given enough data and computational power, an AI could “simulate” wisdom.  This already happens crudely today in therapeutic chatbots.  In situations when there are not enough therapists to serve all clients, an AI chatbot could be used to help people.  These bots work by looking at transcripts of millions of therapy sessions.  If the patient presents with ‘X’ complaint, the bot knows that ‘Y’ answer has had good results in the past.  However, Roshi noted that the bot itself has never really suffered the problem which the patient is presenting with.  It is a useful simulation, but is it really wisdom?


Roshi stopped here and brought up an interesting point.  An AI cannot choose to be a Bodhisattva!  The computer has not experienced the first noble truth and it never can.  It does not have a body which feels pain or a mind which experiences the existential dread of mortality.  The Bodhisattva mind arises because we humans have experienced these things and we see this in other sentient beings.  Because of this we generate compassion and through this compassion we generate true wisdom.  Roshi reminded us that the advent of AI clearly shows us our humanity by what AI essentially lacks.  Because we suffer, we come to practice.  We come to practice to wrestle with the irony of existence and once we have mastered it, we commit to spread this wisdom to all beings.  An AI can simulate this, and someday very skillfully.  However, there is no karma in a computer, and thus no deep compassion, and thus no deep wisdom.  That is the realm of humanity.  


Roshi wrapped up by wishing everyone the happiest of holidays and a happy new year.