The Truth in the Mirror
AI Wisdom and the Human Dilemma
Something Does Not Quite Fit
There is something that does not quite fit in the way we are talking about artificial intelligence.
We speak increasingly of AI as though it were an independent life form moving among us. We ask what it wants, what it will decide, whether it will surpass us, whether it will control us and, at the furthest edge of the argument, whether it might eventually eliminate the human race.
Some of these concerns are serious. Artificial intelligence may acquire capacities whose consequences human beings cannot adequately predict or control. That possibility deserves attention.
But something about the framing continues to trouble me.
Perhaps we are looking in the wrong direction.
What if AI is not primarily the thing we should be looking at?
What if it is a mirror?
And what if the image appearing in that mirror is ourselves?
What Are We Frightened AI Will Become
We are warned that artificial intelligence may become extraordinarily powerful: that it may pursue objectives without compassion; manipulate human beings; value efficiency above human life; and acquire enormous power without a moral compass.
But what exactly are we afraid artificial intelligence will become that human beings have not already demonstrated themselves capable of being?
War did not begin with artificial intelligence. Neither did genocide, slavery, colonialism, torture, propaganda, exploitation or environmental destruction.
AI did not teach us how to turn another human being into an abstraction - an enemy, an inferior, a threat, a statistic, a target, collateral damage.
We learned to do that ourselves.
Perhaps, then, our fear of AI contains another fear beneath it: a fear of ourselves.
And perhaps placing the danger in the machine allows us to avoid looking too closely into the mirror.
The Displacement of Responsibility
There is something comforting about placing danger outside ourselves. If AI is the approaching catastrophe, then it is something being done to us. We become the threatened ones.
In that telling, the human choices that created, financed and deployed the technology slip quietly from view. Responsibility is no longer heard.
But human beings created artificial intelligence. Human beings finance it. Human beings determine its objectives. Human beings decide where it is deployed. Human beings own the companies developing it. Governments recognise its strategic possibilities. Militaries recognise its military possibilities. Investors recognise its potential for enormous profit.
At least for now, behind the machine we continue to find people.
And yet our language increasingly makes those people disappear.
The algorithm decided.
The system identified.
The AI selected.
The model predicted.
Responsibility quietly moves from the human being to the system.
This is not entirely new. Human beings have always been remarkably inventive in constructing systems to which responsibility can be surrendered.
The king commanded it.
The government ordered it.
The market demanded it.
The rules required it.
The computer decided it.
And now: AI did it.
The technology changes. The human dilemma remains.
Who Is Holding the Strings
The most powerful AI systems have not emerged independently of the human world. They are being developed predominantly within enormously powerful corporations and against the background of geopolitical competition between states.
This does not mean that everybody developing artificial intelligence is motivated by greed or domination. Scientists can be motivated by discovery. Engineers can genuinely wish to create something beneficial. AI may contribute enormously to medicine, education, science and human flourishing.
But technology does not exist outside economic and political systems.
Profit matters.
Power matters.
National advantage matters.
Military capability matters.
Prestige matters.
And then comes a particularly dangerous sentence:
If we don’t build it, somebody else will.
Choice begins to disguise itself as inevitability. And responsibility disappears again.
Who Tells Us What AI Means
There is another paradox. Much of what we understand about artificial intelligence reaches us through an information environment increasingly shaped by technological corporations themselves - through search, social media, recommendation systems, advertising infrastructures and now artificial intelligence.
This does not require conspiracy. It is structural.
Even criticism of artificial intelligence can therefore occur inside a framework substantially shaped by the industry developing it.
We are invited to ask whether AI will become conscious, superintelligent, take our jobs, escape human control or eventually destroy humanity. These are legitimate questions.
But while we stare towards that hypothetical future, other questions remain quietly in the present.
Who possesses the power now?
Who owns the technology and controls the infrastructure?
Who possesses the data, trains the systems and determines what they are permitted to see and say?
Who profits and determines which dangers receive our attention?
Who has the power to say no?
Perhaps we are becoming so preoccupied with what artificial intelligence might do to humanity tomorrow that we risk failing to examine what human beings are doing with artificial intelligence today.
Witnessing and the Power of the Image
The same question reaches beyond AI. Who tells us what is happening in the world, and whose account is allowed to become real?
Journalists and media outside the mainstream often keep bringing forward events that larger organisations report later, further down the news cycle or not at all. They warn us, show us and keep the evidence present when power would prefer it to disappear.
Gaza has made the danger of witnessing impossible to ignore. More than two hundred Palestinian journalists and media workers have been killed since October 2023. The Committee to Protect Journalists describes the conflict as the deadliest it has ever recorded for the press. Precise totals vary according to definitions and verification, but the meaning does not: those who have carried images, voices and facts out of Gaza have done so at an extraordinary cost.
Press-freedom organisations have gone further than describing these deaths as the incidental consequence of war. Reporters Without Borders says it has evidence that scores of the journalists killed by Israeli forces were targeted or killed while doing their work, and argues that official accusations and smear campaigns have helped make attacks on reporters appear legitimate. Israeli authorities have disputed allegations in individual cases and have sometimes asserted, often without evidence accepted by press-freedom groups, that particular journalists were affiliated with armed organisations. What is at stake is not only the death of those individuals, but the removal of witnesses from a place largely closed to foreign reporters.
AI-assisted systems also belong within this account. Investigations describe the Israeli military using AI tools to help generate and recommend targets at unprecedented speed. The public evidence does not establish that any system selected journalists because they were journalists. But it does establish that AI-assisted targeting operated within the campaign in which they were killed - and that distinction should not become another hiding place for responsibility. A system may produce a target; a human being designs its criteria, accepts or rejects its output, and authorises the strike.
Information becomes especially dangerous to power when it acquires a human face. Facts and logical argument may be available and still fail to enter us. A photograph, a voice or the story of one family can cross the distance created by official language. It can slow us down. It can make us feel what the abstraction concealed. And in that pause, we may begin to question.
This is why control of images, platforms and visibility matters. Power does not only try to determine what may be said. It also tries to determine what can be seen, whose suffering becomes visible and how long it remains before us.
Warnings From Within
The latest warnings are not coming only from people outside AI development. In September 2026, Jacob Coxon resigned from Anthropic and left the frontier AI industry. Coxon, who had previously worked at OpenAI, warned that companies racing towards self-improving superintelligence were gambling with human lives. His former colleague Evan Hubinger publicly affirmed the seriousness of that concern. These are contested forecasts, not established facts, but they come from researchers close to the systems and cannot honestly be dismissed as ignorance of the technology.
Coxon’s resignation is important not only because he warned. He acted. He chose to withdraw his participation and accepted the consequence of doing so. In a system that presents technological acceleration as inevitable, refusal reveals that the race is still made of human choices.
In July 2026, more than a thousand employees of frontier AI companies also signed the Pacing the Frontier statement, asking for international mechanisms that could slow automated AI development if progress outran safety and governance. There is something both frightening and quietly hopeful here. Many of the younger people building these systems are not morally asleep. They are calling attention to the danger from inside the process itself.
Intelligence Is Not Wisdom
Here we encounter what may be the central confusion in the entire discussion. Intelligence and wisdom are not the same thing.
Artificial intelligence can analyse enormous quantities of information. It can recognise patterns, predict, calculate, compare, synthesise and generate. It can already exceed an individual human being in certain kinds of intellectual task. But none of this establishes wisdom.
Joseph Weizenbaum understood this problem almost fifty years ago. He distinguished between deciding and choosing. Calculation can produce a decision. But choosing involves something more: judgement, values and responsibility.
The central question therefore isn’t simply: Can a computer do this?
It is: Should this be delegated to a computer?
This is an extraordinary distinction because technological culture continually encourages us to confuse the two. If something can be automated, why not automate it? If something can be calculated more accurately by a machine, why should a human being decide? If an algorithm predicts an outcome better than we can, why not follow it?
But perhaps wisdom begins precisely where calculation reaches its boundary.
Intelligence can calculate the possibilities. Wisdom emerges from how we live their meaning.
In 1983, Weizenbaum described the computer as ‘the ultimate technological fix which insulates us from having to look at problems.’ More than a decade earlier, he had warned that public beliefs about computers were shaped by propaganda from the computer community, amplified by the press. He was not merely criticising machines. He was identifying how technological stories can conceal the social and institutional problems beneath them.
Norbert Wiener expressed the boundary in another way: ‘Render unto man the things which are man’s and unto the computer the things which are the computer’s.’ The ethical task is not to deny what computation can do. It is to recognise what must remain ours.
Trust Is Not Reliability
Trust is another human capacity that risks becoming confused with technological performance.
We can rely on a machine to perform a task accurately. Reliability can be tested, measured and compared. Trust is something more. It involves vulnerability. We place something that matters - our safety, privacy, wellbeing or future - into another’s hands.
Trust also has a felt quality. We sense whether words and actions cohere, whether another person is present with us, and whether they recognise that something of value has been placed in their care. Trust develops within relationship and can be deepened, broken or repaired.
Artificial intelligence can produce the language of attentiveness, concern and understanding. It may respond in ways that evoke a genuine feeling of trust in the person using it. But the feeling occurs within the human being. It does not establish that the system experiences the relationship in return.
AI does not presently care that we have trusted it. It cannot feel the consequences of betraying that trust, accept moral responsibility or seek to repair what has been harmed.
This does not make AI useless. It means we must be precise about what we are doing. We may rely on AI for particular purposes while continuing to check what it offers against evidence, lived experiencing and the consequences for others.
As the philosopher Onora O’Neill has argued, what we usually owe an explanation of is not trust but trustworthiness - evidence of competence, honesty and reliability that could justify trust, rather than the feeling of trust itself.
The danger begins when warranted reliance on a tool becomes moral trust in an apparent authority.
And once again, language can make the human actors disappear:
The system can be trusted.
By whom?
For what?
On whose evidence?
Within what limits?
And who will answer when that trust is misplaced?
Perhaps the key sentence is:
AI may provide information we can use. But trust and responsibility must remain with the human beings who choose how it is used.
This also makes the Three Compasses more active: the Inner Compass asks, What in me is trusting this, and why? The Relational Compass asks, Is there genuine reciprocity and accountability here? The Collective Compass asks, Who is asking for our trust, who benefits from it, and who bears the consequences when it fails?
Experiencing
As with trust, so with meaning more broadly: AI’s fluency with the language of a thing is not the thing itself.
This is where Eugene Gendlin offers something particularly important.
Feeling without symbolization is blind; symbolization without feeling is empty.
Eugene T. Gendlin, Experiencing and the Creation of Meaning
Meaning does not exist simply in the symbol. Nor does it exist in feeling left entirely without words. The felt sense is not an undifferentiated feeling. It is an implicit bodily sense of the whole situation: already intricate, not yet fully formed in language and still forming. Meaning can move forward as words and felt experiencing respond to one another.
This is why the problem of colonised language also belongs here: language that manages reality rather than meets it. AI can produce extraordinarily coherent formulations, but coherence of language is not the same thing as moral truth.
Experiencing is always more intricate than any formulation of it.
Eugene T. Gendlin
No formulation - human, ideological, institutional or computational - should be mistaken for the living situation itself.
That distinction becomes extraordinary when considered alongside artificial intelligence.
AI has acquired an immense capacity for symbolisation. It can describe grief, analyse injustice, compare theories of morality, construct arguments concerning human rights, recognise contradictions and produce sophisticated language about compassion. It may even generate words that deeply move the human being reading them.
But what is happening on the other side of those words?
Does grief matter to it?
Can injustice become unbearable to it?
Can another person’s suffering enter its life and change what happens next?
We should be careful here. We do not know what future artificial systems may become. But we have no good reason to equate present AI’s manipulation of symbols with human experiencing.
A vast accumulation of information is not necessarily experiencing. A trillion words do not become a felt sense merely because they are processed together.
Research published in 2025 brought Gendlin’s felt sense into interaction with large language models. Participants worked with AI-generated haiku while checking words against the body for what ‘felt right.’ This suggests a more subtle possibility than either celebrating or dismissing AI: AI can participate in a carrying-forward without being the source of the experiencing that is carried forward.
That is also what has happened in the writing of this essay. AI has been very useful. It has helped me check facts, find further sources that strengthen or challenge the discussion, and offer sentences that might express what I am trying to say - though more often than not, what it offers are general words that cannot hold my felt sense or its implicit meaning. But even the wrong word is not wasted: it points me toward what does fit, toward what feels right, precisely by not being that.
But each formulation still has to be checked against my own experiencing. Is this really what I want to say? Does it carry the essence of what I am trying to make explicit? Does something remain missing, overstated or not quite right? I may sense, ‘Yes, that is close - but there is something else,’ or, ‘No, that does not quite fit.’ Then another formulation can be tried and checked again.
This takes time and attention. The fluency of an answer cannot replace that human pause. Something happens in the interaction, but the felt sense against which the words are checked remains mine. AI can help words to come; it cannot relieve me of the choice to accept them or the responsibility for what I finally say.
Experiencing creates the possibility of wisdom. It does not guarantee it.
A human being is not automatically wise merely because we experience. We can encounter another person’s suffering and turn away. We can feel that something is wrong and rationalise it. Often the obstacle is fear: fear of what acknowledging it might require, what we might lose or what consequences might follow. Courage is the capacity to stay with what is there long enough for a freer response to become possible.
AI therefore exposes a double problem: the machine may lack the experiencing from which human wisdom can emerge, while human beings possess experiencing and nevertheless repeatedly fail to allow wisdom to emerge from it.
Fear and Danger
Fear and danger are not the same. Danger refers to a threat in the situation. Fear is a human response to danger, but it can also be awakened by anticipation, memory or suggestion. It can persist when no immediate threat is present.
The bodily responses that help us survive danger can therefore be activated through communication. Repeated warnings, images, predictions and stories can make threat feel continuous. Fear then enters everyday life quietly. It narrows attention, hastens reaction and makes certainty or obedience feel safer than pausing to question.
AI is highly effective at this too. It can identify what captures attention, tailor messages to particular groups, produce persuasive language and images and repeat them at immense speed. The same technology that can help evidence reach us can also saturate us with fear.
The moral question returns to human choice. Who is using the technology? What response are they trying to produce? Who benefits when fear replaces reflection?
The Curious Word It
And here another small paradox appears. We continually refer to artificial intelligence as ‘it.’
It thinks. It wants. It decided. It learned. It made a mistake.
The word is strangely familiar to anyone who knows Focusing. In Focusing, we also say:
There is something there.
It feels something like...
It isn’t quite that.
It needs some time.
But the word ‘it’ functions very differently.
In Focusing, ‘it’ lets us acknowledge something before it has formed sufficiently to be named. We are curious and interested in what it is, without naming it too quickly or forcing our existing concepts upon it. We remain beside something implicit and allow it time.
We try a word. Does that fit? No. Not quite. Something more is there. We wait.
Perhaps another word comes. And suddenly something shifts.
So in Focusing, ‘it’ keeps the question open while meaning forms.
But in artificial-intelligence discourse something almost opposite can happen. ‘It’ can conceal the already-known human actors by leaving them unnamed.
It decided. Who designed what counted as a successful decision?
It selected. Who determined the criteria?
It predicted. Who decided what should happen because of that prediction?
It learned. From whose world, language, history and values?
The same tiny word that allows an emerging human meaning to remain open can also conceal an existing human responsibility. There is something living in that paradox.
The Felt Sense and the Answer
Focusing does something profoundly countercultural. It allows us not to know yet.
The felt sense is not simply missing information waiting to be retrieved. It is an implicit bodily knowing of a whole situation that has not yet formed sufficiently to be said.
We stay with it.
Something comes.
We check it back.
Does this fit?
Sometimes the answer is: No. Not quite.
That tiny ‘not quite’ matters. It may contain more wisdom than an enormous quantity of perfectly organised information.
AI, by contrast, is extraordinarily good at producing answers. And perhaps one of its dangers is not that its answers will always be wrong. It is that they may become so fluent, immediate and convincing that we gradually lose our capacity to remain with the question.
Perhaps wisdom begins precisely there: where intelligence can no longer simply provide the answer; where we must remain with what does not fit; where we encounter the consequences for another human being; where calculation reaches its boundary; where something not yet formulated needs time; and where ultimately we must accept responsibility for what we choose next.
The Revolutionary Pause
Mary Hendricks-Gendlin called this the revolutionary pause: an interruption of the ongoing situation that creates space for a felt sense to form. Viktor Frankl located human freedom in the possibility of choosing our response. Gendlin locates new possibility in the pause before inherited meanings, cultural roles and automatic reactions take over.
What happens when we begin outsourcing that pause itself?
An algorithm increasingly tells us what to read, what matters, whom to fear, which decision appears optimal, what diagnosis fits, what military target qualifies and what risk is acceptable. The danger is not only that AI might someday acquire too much agency. The more immediate danger may be that we exercise too little of ours.
The pause is not withdrawal from action. It is the place where something not yet named can become a choice for which we are able to answer.
The Moral Compass
Moral orientation can no longer depend solely on institutions, ideologies, technologies or inherited certainties.
This brings us back to our moral compass. Much contemporary AI discussion concerns alignment. How can we align artificial intelligence with human values?
But which human values?
The values we proclaim, or those we practise?
Human rights, government interests, corporate interests, shareholder interests or national interests?
The values of those who possess the technology, or those whose lives will be affected by it?
There is something extraordinary about humanity proposing to align machines with human values while human beings remain so profoundly conflicted about their own.
We say human life has equal value. Our actions repeatedly suggest otherwise. We establish international law and struggle to apply it consistently. We construct institutions intended to protect human beings and watch them become powerless when powerful interests intervene.
We know. And somehow we do not know.
We see. And somehow we do not see.
We feel. And somehow we learn not to feel.
AI did not create this contradiction. It inherited it from us.
Recent philosophical work sharpens this point. Andrej Zwitter argues that treating alignment as though a machine could simply be made to adopt human ethics misses the embodied origins of moral agency: vulnerability, affect, care, conscience, relationship and a lived history. The ethical burden therefore remains human and institutional. The question is not only whether the machine is aligned with us, but whether we are aligned with the values we claim.
A compass does not remove uncertainty.
That matters because one of AI’s strongest temptations is the promise of certainty through prediction, classification and optimisation. Moral life cannot be reduced to optimisation. Sometimes the ethically necessary response begins when a human being refuses the apparently efficient answer and stays with the complexity and consequences of the situation.
What Is True Is Already So
What is true is already so.
Eugene T. Gendlin
It is a deceptively simple sentence. Acknowledging reality does not create it. Refusing to acknowledge reality does not remove it.
And this may be precisely where the AI debate needs to turn towards the human being.
We fear that artificial intelligence may become cold, calculating, immensely powerful and detached from human suffering. Human institutions are already capable of being cold, calculating, immensely powerful and detached from human suffering.
We fear that AI may manipulate truth. Human beings already manipulate truth.
We fear that AI will treat human beings as objects. Human systems already do.
We fear AI acquiring power without a moral compass.
And the mirror asks: Where is ours?
Not as an accusation. As a question. Perhaps as an invitation.
Powerlessness and Turning Away
Perhaps there is another part of this that we do not discuss enough. Human beings today encounter suffering on an unprecedented scale. Every day our screens show us war, displacement, hunger, violence and injustice.
We can know about another human being’s suffering almost instantaneously. But the ordinary individual frequently has almost no corresponding power to change what is happening. There is an enormous tension between what we can see and what we can do.
Perhaps we engage. Perhaps we protest. Perhaps we help one person. Perhaps we despair. Perhaps we become numb. And perhaps sometimes we find another problem.
AI provides an extraordinary elsewhere. We can regulate it, hold conferences about it, calculate its risks, construct safety frameworks, imagine catastrophic futures and identify something that needs controlling.
Perhaps this is easier than remaining with a much older and more painful question:
Why are human beings capable of knowing that something is intolerable and continuing to tolerate it?
There is no technological solution to that question.
AI Cannot Take Responsibility for Us
There is something seductive about imagining that sufficiently intelligent systems might eventually solve the problems humanity has failed to solve. Perhaps AI will help address climate change, disease, economic organisation and conflict. Some of this may indeed happen.
But knowledge cannot by itself tell us what we ought to value. Prediction cannot determine what is just. Efficiency cannot tell us what kind of world is worth living in.
AI may eventually tell us with extraordinary accuracy what is likely to happen if we choose A rather than B. But responsibility for choosing cannot simply disappear into the calculation.
Otherwise we have not solved the moral problem. We have outsourced it.
And perhaps that is precisely what we wanted to do.
The Human Dilemma
Perhaps, then, this was never fundamentally an AI problem. It is a human dilemma. It always has been.
What do human beings do with power?
What happens when our capacity to act becomes greater than our capacity to experience the consequences of our actions?
What happens when intelligence becomes separated from wisdom?
What happens when another human being no longer feels real to us?
What happens when profit becomes more important than life?
What happens when responsibility becomes distributed through institutions until nobody experiences themselves as responsible?
What happens when we create a technology capable of amplifying all of this beyond anything previously imaginable?
And Still There Is Another Paradox
It would be easy to end with a condemnation of humanity. But that would repeat exactly the simplification we are questioning.
Human beings create extraordinary cruelty. And extraordinary compassion.
We exploit. And we sacrifice ourselves for strangers.
We turn away. And we bear witness.
We destroy. And we repair.
Both possibilities belong to us. Perhaps that is the deepest paradox.
The human being capable of ignoring another person’s suffering is also capable of allowing that person’s life to enter their own and being changed by it.
So the question cannot simply be: How do we prevent AI becoming like the worst of humanity?
It must also be: What allows the best possibilities within humanity to come forward?
And: What prevents them?
What Responsibility Requires
The practical task is not to ask AI to become our moral compass. It is to keep human responsibility visible wherever AI is used. That means naming the people and institutions behind consequential decisions; refusing to permit ‘the system decided’ to become an alibi; preserving the human capacity to pause, question and refuse; requiring those who hold power to answer to those who bear its consequences; and attending to present harms while also preparing for future risks.
The Three Compasses become practical when they are asked as questions in the moment of choice:
Inner: What am I sensing? What does not fit?
Relational: Who is before me, and what is happening between us?
Collective: Who holds power, who bears the consequences, and who remains accountable?
These questions do not produce certainty. They keep us in relationship with the living situation long enough for a responsible next step to emerge.
AI can inform all three. It must not become the compass.
AI can help us see patterns and possibilities. It can assist thought. It can even offer words that carry something forward. But the point at which another life may be altered cannot be allowed to become morally ownerless.
The final act remains human: to look, to choose and to answer for the choice.
Looking Into the Mirror
Perhaps artificial intelligence is giving humanity an unexpected opportunity. Not because AI will tell us what is right. Not because it will solve the human dilemma. And certainly not because responsibility can safely be handed to it.
But because the questions we ask about artificial intelligence return with extraordinary precision to ourselves.
What is intelligence? What is wisdom? What is responsibility? What is truth?
What happens when power becomes concentrated?
What does it mean to experience another human being?
What makes something matter? What is a moral compass? And what are we aligned with?
Perhaps AI cannot answer these questions for us. Perhaps neither can we finally answer them. But there is a profound difference between having no final answer and ceasing to ask.
Perhaps one of the most important things about being human is precisely our capacity to remain beside something for which we do not yet have an answer.
What is not yet formed can nevertheless be felt.
Eugene T. Gendlin
To notice: No. That doesn’t quite fit.
To give it time.
To allow what has not yet formed to come.
To experience the consequences for another human being.
And finally to say:
I choose.
Not: The system made me.
Not: The algorithm decided.
Not: AI told me.
But: I choose - and therefore I am responsible.
The crisis of artificial intelligence may ultimately reveal itself as something much older: a crisis of human responsibility.
AI is not simply something humanity has created. It may be becoming a mirror in which we are being asked to encounter what we have created of ourselves.
And perhaps the distinction we most urgently need is not between human intelligence and artificial intelligence. It is between intelligence and wisdom.
Artificial intelligence may become extraordinarily good at answering our questions. Human wisdom may depend upon our capacity to know when an answer is not enough.
The mirror cannot tell us who we are. It can only return the image.
We still have to look.
And looking is uncomfortable because it requires something blaming the mirror never will: accepting responsibility for what we do - and for what we do not do.
Perhaps that is why it is so tempting to look elsewhere. Perhaps blaming the mirror is easier than looking into it.
What is true is already so.
Eugene T. Gendlin
And the mirror is still there.
References and Further Reading
Jacob Coxon resignation and warnings. Ars Technica, 9 September 2026.
Committee to Protect Journalists. Israel Gaza War Attacks on Journalists Across the Region Since October 2023.
Reporters Without Borders. Gazawood The Deadly Accusations Against Gaza’s Journalists, 15 September 2025.
Human Rights Watch. Questions and Answers Israeli Military’s Use of Digital Tools in Gaza, 10 September 2024.
Yuval Abraham. Lavender The AI Machine Directing Israel’s Bombing Spree in Gaza. +972 Magazine, 3 April 2024.
Pacing the Frontier. Statement by employees of frontier AI companies, July 2026.
Anthropic. Responsible Scaling Policy, updated 14 August 2026.
Eugene T. Gendlin. Experiencing and the Creation of Meaning. Free Press, 1962.
Eugene T. Gendlin. Focusing. Everest House, 1978.
Mary Hendricks-Gendlin. Focusing as a Force for Peace The Revolutionary Pause. Keynote address, 2003.
Viktor E. Frankl. Man’s Search for Meaning.
Joseph Weizenbaum. On the Impact of the Computer on Society. Science 176, 1972.
Joseph Weizenbaum. The Computer in Your Future. The New York Review of Books, 27 October 1983.
Joseph Weizenbaum. Computer Power and Human Reason From Judgment to Calculation. W H Freeman, 1976.
Norbert Wiener. God and Golem Inc A Comment on Certain Points Where Cybernetics Impinges on Religion. MIT Press, 1964.
Andrej Zwitter. The Embodied Ethics Alignment Problem of AI. Discover Artificial Intelligence, vol. 6, article 852, 2026.
Onora O’Neill. A Question of Trust: The BBC Reith Lectures 2002. Cambridge University Press, 2002.
Claudia Núñez-Pacheco, Pedro Sanches and Jorge Olivares-Retamal. Searching for the Words that ‘Feel Right’: Resonating with our Bodies and Felt Senses Through Haiku and Large Language Models (LLMs). Proceedings of the 2025 ACM Designing Interactive Systems Conference (DIS ’25).
Patricia Foster. The Three Compasses A Way of Being with Situations.
About the Author
Patricia Foster is a Coordinator, Focusing Trainer and Therapist with The International Focusing Institute and an EAP Certified Person Centred Therapist. Her work brings experiential philosophy into Focusing training and Person Centred and Focusing Oriented therapy.
More of Patricia’s writing: www.focusinglifeskill.com
In memory of Maher Al-Afifi, journalist and director, who spent years documenting the suffering of others with dignity and humanity — and in support of his wife Reham and their four children, Roa, Mahmoud, Fatima and Lina, who now carry what his death left behind.
Comments