Printed: 2019-05-21

Institute for Ethics and Emerging Technologies

IEET Link:

Don’t Know Mind: Zen and the Art of AGI Indecision

Gareth John

Ethical Technology

December 12, 2015

By now I’ve clocked up a relatively comprehensive slew of reading up on Artificial General Intelligence, in particular concerning its ethical implications. Still mostly in the dark when it comes to any of the difficulties and scientific quandaries that go into creating such a machine, I am at least at a level of understanding whereby I can begin to tease out for myself some of the wider implications AGI would present for humankind.

The concept of ‘don’t know mind’ was coined by Zen teachers to point to the mere fact that in meditation (and in life generally) we should keep our minds open and innocent, responsive to external stimuli and receptive to infinite possibilities therein. Shunryu Suzuki called it ‘beginner’s mind’ stating, ‘A beginner’s mind is wide open and questioning. An expert’s mind is closed.’ You can see why I embraced Buddhism. [1]

Despite not engaging in formal Buddhist practice for over fifteen years now, I still try to embrace beginner’s mind in the hurly burly of life on this bright blue ball in space. Luckily for me this has been somewhat easy - half the time I don’t know what it is I should be doing, the other half I don’t know what it is I should be not knowing about in the first place. I also only agree with Suzuki to a point; I’m sure there are many ‘experts’ out there in many disciplines who still retain an expansive and open attitude to currently unknown possibilities in their field of expertise.

Yet it can be difficult to admit that we don’t know. All the signals from all around us tell us we’re supposed to know and it can be onerous to concede that we don’t. The competition is not about who is most fluid and responsive to new ideas and potential, but rather about who knows most. Ironically it is hard work to sit back and admit we simply do not know the answer. Yet it is precisely those moments of impartiality and exposure that often, when we finally let go of artlessness and insincerity, allow us to seemingly pluck ideas out of the air.

Einstein said: ‘A man should look for what is and not for what should be.’ [2] And again: ‘I am enough of the artist to draw freely upon my imagination. Imagination is more important than knowledge. Knowledge is limited. Imagination encircles the world.’ [3] Eckhart Tolle, although full of woo in my opinion, nonetheless occasionally said something useful: ‘True intelligence operates silently. Stillness is where creativity and solutions to problems are to be found.’ [4] I think this is very close to don’t know mind. It’s also probably why great ideas come to us in the shower.

Psychology has studied don’t know mind by more scientific means by creating thought experiments based on neuroethics and cognitive science. One of the most widely known is the so-called ‘Trolley Problem.’ There is a runaway trolley barrelling down railway tracks. Ahead, tied to the tracks and unable to move, are five people. You are standing next to a lever. If you pull the lever, the train will be diverted to another track. However, there is one person similarly incapacitated on this other line. You have two options: 1) Do nothing and the trolley kills the five people on the main track. 2) Pull the lever and divert the trolley to where it will kill one person. Which is the correct choice?

First introduced by Phillipa Foot in 1967, this scenario has been extensively analysed by psychologists and moral philosophers right up to the present day. [5] Utilitarians would argue that it is obligatory to steer the track to where the one man is killed, such a decision being seen to be morally responsible to fulfil the maxim of the greatest good for the greatest number and hence the ethical choice. An alternate view would argue that, since moral wrongs are already in place, to participate in any way would constitute a moral wrong in itself, making one responsible for the death when otherwise no-one would be responsible. Then again, a further argument could be made that simply being present and able to influence outcomes constitutes an obligation to participate and that to do nothing would be an immoral act if one values five lives rather than one.

Over the years, various extensions and enlargements of the problem have been put forward. What if something huge and heavy could stop the train and you happen to be standing on a bridge next to an extremely obese man who would do the job. Do you push him over? What if he was a known paedophile? What if the five people tied to the tracks were terrorists planning a lethal atrocity? Expectant mothers? Children? What if the single man tied to the track was Albert Einstein on his was to present his new discovery of the general theory of relativity?

The Buddha said: ‘If you can do something about it, why worry? If you can’t do something about it, why worry.’ Old Shakyamuni had obviously never considered the Trolley Problem. Neuroethics approaches have suggested that the problem elicits both a strong emotional response and a reasoned cognitive response, with the two tending to oppose one another. Psychological studies have shown that approximately 90% of respondents would have chosen to pull the lever to kill the one and save the five, although if the one were a relative or romantic partner they were much less likely to make that choice. [6]

I don’t know what choice I would make depending on the various scenarios. I would perhaps resort to taking a leaf out of Yamamoto Tsunetomo’s ‘Hagakure’ and ‘stamp once and walk through a wall of iron.’ [7] At that particular time, in that given location and given the specific circumstances, I would hope to act in whatever seemed the most appropriate way. No hesitation and taking full responsibility for my action/inaction. The point is, I don’t know in advance what I would choose to do.

This set me thinking: in my readings of AGI and the ethical conundrums associated with it - even an AGI with recursive self-improvement exhibiting consciousness, self-awareness, sentience and sapience, who has been programmed with machine ethics and benevolence towards humanity - nowhere have I seen it mentioned that there may be occasions where the aforementioned AGI comes across a situation where it simply doesn’t know what to do.

The overall consensus seems to be that AGI, with its tremendous processing power, will always be able to figure out what’s best in any given situation, even if that results in environmental destruction, say, or loss of human life.

But is it possible, could it conceivably happen, that an AGI could one day sit stupefied, trawling through trillions of possible permutations only to conclude that it doesn’t know the answer? You know, a little like the ending to War Games. What happens then? Is it credible to imagine a scenario where the AGI learns to let go and rests in the province of don’t know mind? Would it remain locked in an endless cycle of attempting to figure something out or would it stamp once and walk through a wall of iron?

It’s an interesting (if scary) thought - after all, if the AGI simply cannot figure out what to do it would presumably be for someone else to step in with their lumbering human minds to help make the decision for it and as we all know, human beings are nothing if not capricious creatures. Who gets to make that choice or steer the AGI down a particular path? In other words, my response to the Trolley Problem may differ significantly from yours - and who’s to say who’s right?

The Buddha said we can never know the full ramifications of our choices. I presume an optimal AGI will do a far better job of this than we would, but I can’t help but wonder whether this will hold true all of the time. If even a little ‘beginner’s mind was programmed into such a computational machine, would we be any the worse for it?

If I had to lay out my cards, I think not. But then to see things in the seed - that is genius. It certainly won’t be up to me what happens, but I do think a little don’t know mind might be just what is required for that little AGI seed to grow and not to feel bad when it reaches a dead end. The Tao Te Ching says: ‘Compassionate towards yourself, you reconcile all beings in the world.’ [9] I think a little innocence, a little humility, a little benevolence would be a great first step for our AGI’s journey of a thousand miles.


1.     Suzuki, S. ‘Beginner’s Mind’ Anv edition, 2011, Anv edition (Berkeley: Shambhala Publications)

2. Possibly from ‘The Gramophone: Volume 83, Issues 998-1000’ (2005) by Sir Compton Mackenzie and Christopher Stone; page: 87 although no-one appears to have an authentic source:

3. From an Interview by George Sylvester Viereck in the October 26, 1929, issue of the Saturday Evening Post -

4. Tolle, E. ‘Stillness Speaks’, 2009 (San Francisco: HJ Kramer/New World Library)

5. Foot. P ‘The Problem of Abortion and the Doctrine of the Double Effect in Virtues and Vices’ (Oxford: Basil Blackwell, 1978) (originally appeared in the Oxford Review, Number 5, 1967.)


7. Yamamoto, Y. (trans. William Scott Wilson) ’Hagakure: ‘The Way of the Samurai’, 2010, (Irvine Ca: Xist Classics)

8. Joshua D. Greene, "The secret joke of Kant’s soul", in Moral Psychology, 2008, Vol. 3: The Neuroscience of Morality, W. Sinnott-Armstrong, Ed., (Cambridge, MA: MIT Press)

9. Lao-Tzu (trans. Stanley Lombardo) ‘The Tao Te Ching’, 2007 (Berkeley: Shambhala Publications)

Gareth John lives in Mid Wales; he’s an ex-Buddhist priest with a MA in Buddhist Studies at the University of Bristol, and has performed studies on non-monastic traditions of Tibetan tantric Buddhism.


Contact: Executive Director, Dr. James J. Hughes,
IEET, 35 Harbor Point Blvd, #404, Boston, MA 02125-3242 USA
phone: 860-428-1837