AI and Buddhism

Throwback to 2018 and Rinpoche’s words on Buddhism and Artificial Intelligence, which are more relevant now than ever.

BUDDHISM: PATH TO BODHI — PATH BEYOND INTELLIGENCE
Dzongsar Khyentse Rinpoche
Keynote Address at 8th World Youth Buddhist Symposium on Artificial Intelligence and Buddhism, 25th July 2018, Chiang Rai, Thailand

We are here to share our thoughts on what Buddhism can offer in the age of Artificial Intelligence.

But in order to do that, we need to establish a common language, a way to communicate between the worlds of science and Buddhism — if that is even possible. If we can’t agree on basic definitions, then we will just be talking past each other at this conference — intending, meaning and understanding different things, let alone answering the kinds of searching and profound questions this conference has posed. I’ll take just three basic examples:

1. Can science and Buddhism communicate? — “Life”, “mind”, “paradox”
The conference brochure asks: “Is life just data processing”? But what do we mean by “life”? The Oxford English Dictionary defines life as the activity that “precedes death”. But Buddhism does not exclude the continuity of consciousness after death. So, as Buddhists we might ask:

• Will artificial intelligence be subject to continuity?
• Will it be subject to the illusion of time?
• And to put it crudely — will A.I. reincarnate?

For that matter, what do we even mean “intelligence”? If the human mind is nothing but brain activity and biological function, there is no common ground between science and Buddhism, because for Buddhists, mind is something altogether different: It is what cognizes, gets habituated, suffers from hope and fear, and gets so distressed and out of control that it gets trapped in its own illusions. But that mind is also what can be trained and tamed so that it learns how to undo its own delusions. It is not just intellectual but also intuitive, and it has the capacity to be altruistic as well as selfish.

Buddhism and science may have entirely different notions of “knowing”. In fact, modern science is said to be based on the principle of ignoramus — Latin for “we do not know”. It assumes we don’t know everything and that all we think we know can be proven wrong as we gain more knowledge. I admire that creed.

At the same time, we Buddhists have concluded that certain truths — such as all compounded things are impermanent, all emotions are pain, and nothing has truly existing nature — cannot be disproved. So, while the starting point for scientists is “we don’t know”, the starting point for Buddhists is that “we have noble truths that can’t be proven wrong.”

The conference brochure worries about the notion of ‘singularity’ — that technology will escalate out of control to the point where human beings will be outsmarted and overtaken by A.I. machines. But what’s so new? We humans are already very good at creating things that get totally out of our control: we created a God who became uncontrollable, and an Industrial Revolution that also created the plastics that now choke our oceans.

In fact, it’s because we don’t want to suffer out of control like this and because we want to discontinue the dualistic cognition that underlies our suffering, that the Buddha’s teachings on selflessness, dependent arising and shunyata are as potent and relevant now as ever.

Of course Buddhism doesn’t deny the influence of the brain — or of a toe or the weather for that matter. But the vastness, subtlety and complexity of the Buddhist view of “mind” go far beyond any of those causes and conditions. In fact, we could say that all of Buddhism is the study of this mind.

For Buddhists, even the most fundamental search for a reference point or a “knower” is a habit we want to undo. Now if A.I. can handle that and induce non-dual realization, then A.I. is dharma and Buddhism is obsolete.

But from my very limited knowledge of A.I., I doubt it shares that view of mind, and I therefore see no necessity for the essence of Buddhism to change at all. I don’t know how the A.I. revolution will affect monotheistic religions like Christianity and Islam, but I am confident we Buddhists have no cause for concern. In fact I see the four noble truths as even more compelling and relevant a hundred years from now as we become ever more alienated from our true being.

I wonder: can A.I. create a perfect world in which we are never alienated and that overcomes even the human proclivity to alienation? Can it deal with subtleties like notions of individualism and individual rights that are celebrated in western democracies today, but which seem to signify that taste for alienation?

For that matter, I wonder whether artificial intelligence has the power of “belief” — both educated belief and especially blind belief — which is the driving force for our so-called “life” and “happiness”.

Aside from differences in how we define and use words like “life” and “mind”, there are other obstacles to meaningful dialogue between science and Buddhism. For example, while science seems uncomfortable with paradox and seeks to resolve it, the quintessence of Buddhism is profound appreciation of the paradox of everything and therefore to be totally comfortable with it.

Thus, the moment a trained Buddhist mind sees form, it also ideally sees emptiness and thus is freed from blind hope, and the moment it sees the empty nature of reality, it also sees form and is thus liberated from fear.

That’s why we Buddhists won’t conclude that non-apparent phenomena like God, reincarnation or luck don’t exist. Nor do we confirm that there is a head on my neck, a floor under my feet and a ceiling over my head. So, we don’t confirm anything as either truly non-existent or as truly existing.

That’s also why Thai people may learn the truth of egolessness — the nonexistence of the self — and at the same time have no qualms performing meritorious acts like visiting temples, offering flowers, or giving alms to a begging monk.

And so, for Buddhists, pain and suffering arise when form and emptiness are divorced, when existence and non-existence are divorced, when how things are and how they appear are divorced. Fundamentally, we suffer when we fail to appreciate these paradoxes.

So maybe instead of asking what Buddhism can offer in the A.I. age, which may be no different than it has always offered, we could flip the question around and ask if A.I. has any interest in the fundamental issues of concern to Buddhists. It may not be up to Buddhism to conform to science, as we too often implicitly assume, but perhaps time for scientists to grapple with the truths the Buddha taught 2,600 years ago.

I don’t know whether we will ever resolve the differences in definition and view between science and Buddhism sufficiently to create a real dialogue. In fact, I don’t even know whether such a dialogue is more relevant today than one between Buddhism and economics, political science, or poetry. I for one will not classify the latter as less important or worthwhile.

But I am sure that unless we start by acknowledging and exploring such basic differences in view and definition, we won’t be able to answer the searching questions this conference poses.

2. Absurd questions?
To stretch our imaginations enough to grapple with these differences between science and Buddhism in both view and definition, we shouldn’t shy away from questions that might seem absurd:

Suppose for example that I have a robot as a student. He or she is programmed to be calm and without anger, jealousy and emotion. I may be a bit impressed and not prescribe shamatha meditation for such a student. But if I’m a half-decent teacher, what I really want is for this student to stop having the goal of being calm and peaceful, or even being the one doing the calming. After all, a follower of the Buddhadharma isn’t trying to get a medal for being the most calmed or calming person, but rather wants to shrug off all those references of being calm, getting calmer, or being on the path of calmness.

I am no expert on A.I., and my very limited information is from a few articles and hearsay. For example, I’ve been told that 50 years from now, I’ll be able to download myself onto a computer that will be able to think, talk, and respond just like me. But if that device then has all our collective habitual patterns like anxiety, uncertainty, and insecurity, then it’s become just another ignorant sentient being and object of compassion that needs to be enlightened. In that case, the only difference between him and me is that I was born from my mother’s womb while he was born in a laboratory with the help of some other robots.

And what about karma? Is it bad karma to smash the computer or not to charge it? If a computer can programme itself through expert systems and feedback loops, and if it can therefore think and act like a person, does it have good or bad karma if it saves a life or kills someone?

On the face of it, such ‘Buddhist’ questions may seem absurd. But remember that Google has already invented expert systems that developed a new language of exchange that its own inventors couldn’t understand, prompting Google to shut it down. So A.I. experts and policy makers might benefit from Buddhist insights into the nature of reality before they let their systems run the economy, health care and military on the basis of internal workings none of us can understand. Indeed, it is no exaggeration to say they ignore Buddhist wisdom at their, and our, peril!

At a more prosaic level, would I prefer an A.I. girlfriend who says and does exactly what I want and need? And if so, do I want to know if she is an artefact of artificial intelligence or not? Again, it is clear that many of the questions that occupy A.I. people are of little interest to Buddhists who relish dwelling in the reality of dependent arising and human uncertainty.

In other words, it sounds to me as if proponents of artificial intelligence are boasting that their systems will make things more predictable. But if that is so, then there will be no more fun even in our ordinary human world. After all, much as we humans think we hate uncertainty, it is uncertainty that drives the economy, governance, management, and everything else in our lives. Profit derives from uncertainty, and it is this unpredictability that, at an individual level, we cherish as romance, love, and relationship.

I was also told that artificial intelligence may extend our lifespans. But that will still not change the Buddhist wisdom of knowing that all compounded things are impermanent. And I’ve been told that medical systems will better be able to diagnose diseases, which of course is excellent. And yet, as I said, the truth of impermanence remains real, and it is precisely within this impermanence, uncertainty and unpredictability that we find life.

So perhaps it’s time to ask what the purpose of artificial intelligence is. If it is to make systems like medical diagnosis and weather forecasting more accurate and therefore predictable, of course that’s excellent and furthers what human beings have always tried to do.

But is the purpose of artificial intelligence to conquer time and space by making things permanent and unchanging? If A.I. can actually do that, then perhaps dharma is obsolete after all, because going beyond time and space is what we Buddhists call nirvana. But I doubt that’s the motivation of the creators of artificial intelligence as human beings, simply because the goal of all scientific advancement is surely just to improve our human lives, which has nothing to do with enlightenment.

3. Changing times — happier or not?
We get so excited by the latest technology, and whether it can “save or doom humanity” as your conference brochure asks, that we forget how unbelievably life-changing have been the technological revolutions of the past.

The first Agricultural Revolution, that transformed human societies from hunting and gathering to farming, happened thousands of years before the Buddha. While it improved food security, that revolution also led to all the problems of politics, armies, traffic and more that plague us today.

Then the Industrial Revolution gave us radios, televisions, cars, aeroplanes and other conveniences that were inconceivable just 150 years ago, and it also gave us atomic bombs and global warming that can destroy the world. If temperatures and sea levels keep rising as they are, Bangkok will sink under water within 20 years and our grandchildren will see London flooded. If A.I. can stop that, it would be great.

But have all the huge changes in our lives and lifestyles made us happier? I get what I want so much faster. But that speed might increase anxiety because it only takes a second instead of a month to get bad news. Maybe we can get happier faster and also upset faster.

So, through all these changes, and before we get too excited by the new digital and A.I. revolutions, let’s recognize that one thing hasn’t changed — our basic anxiety, our insecurity, our ignorance and our search for happiness.

And so long as that exists, Buddhism has just as much to offer now as 2,600 years ago. In fact, because Buddhism is not bound by time, place, culture or geography, it can incorporate, adapt and speak to whatever the needs of any people and era are — including those of our A.I. age.

12,000 years ago, when there were less than a million human beings on this planet, they lived in constant fear of being eaten. But so long as there is hope and fear — whether we’re in danger of being eaten by tigers or having our lives totally dominated by A.I. 500 years from now — every word of the Buddha’s 84,000 teachings will be relevant.

4. So what can Buddhism offer in this age of artificial intelligence?
If we accept the reality of anxiety and suffering, regardless of time and conditions, then one thing Buddhism can offer is the inseparability of wisdom and skilful means — the wisdom of the right view and the skilful means to recognize that view and maintain the wisdom.

What is the right view? That while things appear, function and seem to continue, not a single thing truly exists. Everything is like a dream, a mirage or a rainbow, and we suffer when we don’t have this view. In fact, this is why Buddhism regards wisdom as higher than morality and ethics, and it’s why Buddhism can answer the challenges of technology in our era of artificial intelligence far more effectively than systems based on morality.

As Buddhists, what we want is to know the truth and to rid ourselves of the habits that obscure and stop us seeing the truth. If an artificial intelligence device can help us do that and provide information that aids our pursuit of the truth, that’s excellent and I’ll gladly use it.

So for example, if A.I. can come up with a device that tracks the neurological channels through which habits work and dig them up and short-circuit them, that’s fantastic. I’ll buy it. But unless artificial intelligence can deal with the fundamental anxiety that has plagued human existence since time immemorial, and which actually defines and embodies what we call human life, then what Buddhism can offer won’t change at all in its essence.
How Buddhists offer that unchanging ancient wisdom, however, will change dramatically:

5. Buddhism won’t change. Buddhists have to change — a lot!
So, on the one hand, A.I. is not going to change Buddhism and its validity and value for our world. On the other hand, Buddhists — especially traditional Buddhists — will have to change enormously. And if we’re to be helpful and relevant to people, our Buddhist world will have to be fully aware of this difference.

Forget even the A.I. revolution: we should start by acknowledging that we Buddhists have not been good at changing the way we do things — of delivering the Buddha’s message to people in the right way, at the right time, and in the right place so that they can understand and practice it.

From Japan to Sri Lanka to China to Thailand to Korea, Buddhists are so stuck in their old traditional cultures that have nothing to do with the essence of Buddhism, that they’re in danger of becoming irrelevant to people in our modern world.

For that matter, even the notion of culture itself has to change. In our globalized age — with jeans, pop music, Hollywood and Bollywood penetrating every corner of our world — the very concepts of nationalism and national culture may have to go.

So, if we as Buddhists are going to be of any use to our world and any help to suffering beings, we have to pry ourselves loose from our own tight cultural habits and convey the Buddha’s message in a way totally relevant to the time, age, conditions, and people we’re addressing.

And yet none of those changes — A.I. included — are going to change the way Buddhism functions in its essence. So long as beings have what Buddhists call mind — but what others might call brain or data or whatever you want to call it — as long as this mind has anxiety, habits, cognizance, likes and dislikes, and is subject to time, and so long as existence is suffering, then Buddhism will be as relevant to the age of A.I. as it was at the time of the Buddha.

So, Buddhism will become obsolete only when duality, ignorance, anxiety, hope and fear cease to exist — which of course is the entire aim of Buddhism. Indeed, it is for this basic reason that Buddhists have never prayed: “May all become Buddhist.” We only pray: “May all become buddha”.