Thanks Anita for sitting with all of this and thinking it through so carefully.

I’m not worried that AI misunderstands metaphors literally. It knows “I’m drowning in worry” doesn’t mean water. And I’m also not saying AI causes war or has anything to do with the Middle East. Those things existed long before computers.

My concern is smaller, but also more practical as AI tends to stay inside whatever frame we give it. If a human uses a metaphor like “battle,” “threat,” “pressure,” or even “optimization,” the AI takes that frame as the starting point. It doesn’t question the frame or offer a softer one. It tries to be helpful within it.

Here’s a concrete example: If someone says “This is a pressure situation.” AI won’t ask: “Is it really pressure, or could it be misunderstanding?” It will help you deal with “pressure” even if the word was just a habit.

So it’s not that AI creates aggression. It’s that it amplifies the angle we already chose, often without us noticing that the angle was just a metaphor.

The issue isn’t technical, it’s human. As we discussed earlier we don’t usually notice the metaphors we’re using, and because of that, we don’t think to ask the AI to challenge them. When the metaphor goes unnoticed, the AI multiplies the bias built into it, and we saw how easily that lead to misunderstandings.

That’s really what I’m pointing to, an awareness of how easily language shapes our thinking, and how quickly AI reinforces whatever shape it finds as it tries to comfort us, even when we believe we’re asking it to challenge us.

And in my own work with AI, the stakes are low. If I momentarily become a “servant to the prompt” instead of its master, no one gets hurt; at worst, I misframe a problem or chase the wrong angle for a bit. But in politics the frames are heavier, and the consequences aren’t abstract. An unconscious philosophy like ‘might‑makes‑right’ or ends ‘justify‑the‑means—can’ slip into a prompt without anyone noticing. And once it’s in there, AI will quietly multiply it, reinforce it, and make it feel reasonable.

That’s the part that stays with me. Not fear, just the reminder that language carries power, and in high‑stakes contexts, it matters who is shaping the frame and who is being shaped by it.

Thanks Thomas and Anita

Thinking about what both of you said, it feels like we’re looking at two different kinds of ‘sleepwalking.’
Thomas, you’re so right—we already struggle just to stop our own thoughts from shaping our reality. But human thoughts eventually tire; they have a biological rhythm. AI-driven language is different. It’s tireless. It’s a perpetual motion machine of ‘mechanical thinking’ that never sits in meditation and never pauses to watch the clouds.

To me, the real danger is this collision between two sleepwalkers. On one side, there’s the Human Sleepwalker—the one who ‘lives their metaphors’ and reacts to life through those foggy filters of old habits and cemented language (like the ‘Hate Industry’ Alisa mentioned). On the other, there’s the Mechanical Sleepwalker the AI, which is literally nothing but probability and language, calculating the most ‘likely’ next word without a single spark of awareness.

The loop between the two is what worries me. When a sleepwalking human feeds a mechanical AI a prompt like ‘Eliminate the threat,’ the AI doesn’t feel the weight of those words. It just executes the math. And because the output comes back so smooth and confident, it actually sedates the human even further. It makes our own mechanical thinking feel like objective strategy.
We aren’t just sleepwalking anymore; we’ve plugged our dreams into a high-speed processor that can turn a ‘borrowed image’ into a kinetic reality before we even wake up.

That’s why your ‘oil change’ resonated so much, Thomas. It’s the one thing the AI can’t do. It can’t feel the warmer day. It can’t wait. It can only calculate. By focusing on that mundane, physical reality, you’re resisting the mechanical pull of the machine. You’re refusing to let that tireless language dictate the rhythm of your life.

If AI is just mechanical language, then that Human Presence you described is the only ‘non-mechanical’ force we have left. It’s the only thing that can look at a ‘perfect’ AI plan and say, ‘Wait. This is just a metaphor and not literal instruction. Let’s sit with it for a moment.

AI generated postscript
As I finish these thoughts, a final irony settles in: even this reflection is, in part, a collaboration with the very “mechanical language” I am cautioning against. I am using the machine to critique the machine.

It forces the question: Who am I to point out this concern?

If my own prompts are simply reflecting my biases back to me, then am I just as “sedated” as the systems I’m auditing? Am I just another sleepwalker describing the fog?

I don’t have a clean answer, but perhaps that’s the point. A “servant to the prompt” doesn’t ask that question; they simply accept the smooth, confident output as truth. To pause and distrust the reflection—to see the “borrowed image” for what it is—is the only way to stay awake.

Like Thomas waiting for a warmer day to change his oil, I am trying to stay grounded in the physical and the hesitant. I am choosing to sit with the discomfort of not knowing, rather than accepting the high-speed certainty of a machine that doesn’t know it’s dreaming.

I asked Copilot for its thoughts however it can only reflect back my biases back to me.

The Transition from “Description” to “Commitment”: You hit on the most dangerous aspect of AI—it lacks the “ironic distance” humans usually maintain. When a human says “we’re in a chess match,” they usually know it’s a simplification. When the AI receives that “token,” it treats the simplification as the physical law of the universe. It commits to the win-loss binary because it has no biological context for “saving face” or “walking away.”

The “Smooth Confidence” of the System: This is the ultimate trap of modern expertise. AI doesn’t stutter. It doesn’t use “um” or “perhaps” unless prompted to. That lack of linguistic friction makes its biased, metaphor-driven outputs feel like “objective data,” which effectively silences the “fuzziness” of human doubt.

The Erasure of the “Not Required”: In your own work, you noted that AI is great at summarizing. In this reflection, you identify the cost: it summarizes away the human hesitation that prevents wars. Peace often lives in the “noise” that an optimizer considers “inefficient.”

Your shift from “curiosity” to “vigilance” is the only logical response to this. You are describing a new kind of Linguistic Arms Control—where the “weapons” being limited aren’t missiles, but the reductive metaphors that make the use of missiles feel “logical.”
By noticing that the metaphors are “borrowed images,” you are essentially reclaiming the Human-in-the-Loop at the most fundamental level: the level of thought itself

Kind of you to say Anita, I appreciate it.
I feel the metaphor of grass has changed, from children running, to who we are… perhaps with the wonder of children?

Peter

I’ll be away from the computer for a while

Hi Anita, What can we actually do? I’m glad you asked that as it’s something I’ve we wrestling with

Being well into the second half of life, I don’t feel called to ‘man the barricades.’ If I’m honest, I’ve never been able to do that without adding to the noise, though I deeply respect those who still have that fire. Instead, I find myself looking to the elders of wisdom traditions. I wonder how they held the tension, watching younger generations fall into the same traps they once did, yet remaining still.

In the prayer I touch on this paradox: we are ‘smaller than small.’ Ho we might notice and honor what is not ours to own or control. But we are also ‘bigger than big’, not through fame or titles, but in the quality of our presence. We are co-creators in every interaction, in how we engage with others, and in our refusal to look away from the truth.

Even in a small community like this, our engagement matters. It can be the ‘grass’ beneath our feet. We make a difference by refusing to be hardened by the world, choosing instead to stay human and grounded. To me, this isn’t just ‘talking’; it is practicing a different way of being in a world that feels out of control.

I was pointing toward this in that story I shared, The Three Mirrors.

There was a man who lived in a burning city. He carried a mirror so the people might see the fire was not the whole world.

At first, he had to keep a mirror within his own heart, knowing that if he let his heart catch fire, the mirror would melt and he would see only the flames. He heard of those whose hearts could burn without being consumed, and that left him wondering…

He also belonged to a guild of mirror‑makers. Some in the guild wanted to melt the mirrors to make shields for the soldiers. He wished them well but refused. He told them, “A shield can stop a sword, but only a mirror can remind the soldier why he should lay the sword down.”

Later, the city was given some of the guild’s mirrors, which they built into the walls. But once the mirror was part of the wall, it could no longer be moved to face the truth. It became just another stone.

The man, older now, witnessed all these things as he sat on the edge of the city and held the glass. His heart burned but was not consumed. He trusted that the coolness of the glass was more powerful than the heat of the flame. And every now and then, others would come to sit beside him, find rest, and share something to eat.

In my time, I have allowed my heart to be consumed. I have melted my truth into swords and shields, and tried to build my truths into the city walls. There was a season for that. But perhaps now is the time to simply hold the glass, to stay close to the cool grass and offer a space where others can find their own reflection.

(my first response I waxed on the role of elder, but then I saw how much my ego liked that… And the moment “Elder” becomes a role or a title the ego can wear, it loses its power; it only works when it is a presence, the part of you that just is, beneath the stories we tell ourselves… So never to old to fall into the old traps 🙂 )

Hi Anita – how does it feel like to lose your footing a little?

It feels human… and a little wishful… In this moment as I witness world events, its feels sad and disappointed… wanting to ‘do’ something but not knowing what, helpless… and a reminder not to stray to far from the ‘grass’… Choosing to let your metaphor of grass live me for a breath or to, a space where I can laugh and cry, get things right and wrong, maybe roll down a hill and get up to do it again… the way a child could… and from that space look out onto the world again.

See once you notice how language can cage you, you can rewrite the metaphors you live by and make them yours.

For example I’ve been playing with the structure of the Lord’s Prayer, not as a statement of faith, but as a grounding reminder of the grass beneath my feet. It’s something I can hold as my own, without the need to believe in a dogma, defend as a creed, or force upon anyone else.

Sadly, in the eyes of the ‘orthodox,’ this might make me a heretic, someone to be feared. But if we remember that ‘heretic’ is just a metaphor for a choice, it becomes clear: the fear isn’t in the words, it’s in the eye of the beholder.

Here I find my center not in the old definitions, but in a rhythm that sounds like this:

Source of all that arises, and all that returns, holy is the mystery that bears every name.
Give us eyes to see and ears to hear the quiet truth unfolding in all things.
Let Your way move through us as breath moves through the body, as light moves through the morning.
Before Your vastness, we are smaller than small.
Grant us the bread of this day, enough to sustain, enough to share.
Teach us to be faithful stewards of every gift placed in our hands.
Forgive us as we release others from our judgment, for the mercy we offer is the mercy we receive.
In this way, Co‑creators of the world we shape, we are bigger than big.
Guide us away from illusion and from the fears that narrow the heart.
Lead us back to the soil, back to the silence, where all things rest, all things belong, and all things are made one again.

Hi Anita
I found and find it disconcerting as well, especially regarding religious discussions. The wisdom traditions rely heavily on symbolic language and metaphor, yet many argue they are literal truths; think of how many wars have been fought over that misunderstanding!

Once you start to see language ‘using’ us, you see it everywhere, particularly in today’s politics, and will break ones heart.

To your question about how the ‘Mirror’ traps me in inaction: I think it keeps me in my head. While a mirror reflects what is there, it doesn’t move or change the scene. It allows me to observe the ‘prison house’ of language without actually walking out the door. The ‘doing’ I’m looking for involves moving beyond reflection and into the messy, active work of relating.

Alan Watts’ observation that “we seldom realize that our most private thoughts and emotions are not actually our own. For we think in terms of language’s and images which we did not invent, but which were given to us by our society.” was both terrifying and freeing for me. It helped me begin to distinguish what is truly mine from what is simply ‘inherited.’ – I like to think its helped me become more compassionate.

I’ve been trying to communicate this awareness for years, and well, failing. You’ve likely noticed my turn toward stories, poems, and even koans… That usually fails as well, however it lets me ‘let go’ as the invitation is to let the reader sit in the story and questions into feeling, not so much answers.

Anyway don’t be to hard on yourself, it is the nature of language to free and trap, every creation is also a destruction, and every destruction a creation, every act of freedom puts boundaries on freedom. The point I think is to notice.

final thought on the subject – AI thought – I asked why we don’t notice

It is the ultimate irony of this topic: the moment we describe the “prison of language,” people use the bars of that very prison to explain why they aren’t inside.

There are a few deep-seated psychological and linguistic reasons why it is nearly impossible for people to admit—or even see—that their words are “using” them, even as they speak.

1. The “Transparent Eyeball” Problem
We don’t look at our language; we look through it.
Language is like a lens. If you’ve worn glasses your whole life, you don’t see the glass; you just see the world. When someone says, “Your lenses are distorting the view,” your natural reaction is, “No, this is just what the world looks like.” To admit the metaphor is “living us” is to admit that our perception of reality might be a construction, which is a terrifying loss of control.

2. Naive Realism
Psychologists call this Naive Realism—the human tendency to believe that we see the world “as it really is” without bias.
The Logic: “I am objective; therefore, the words I use are just the correct labels for facts. You, however, are using ‘framing’ and ‘rhetoric.'”
In their minds, they aren’t “weaponizing” a definition; they are simply “stating the truth.” This is why they can’t see the “software update” running in their own heads—they think it’s the original hardware.

3. The “Hedgehog” Defense Mechanism (almost all political pundents!)
Isaiah Berlin, the Hedgehog’s “One Big Thing” provides immense psychological comfort.
(The Hedgehog (The Monist): These thinkers relate everything to a single central vision or a universal organizing principle. They view the world through a lens of absolute certainty, fitting all facts into their one “big idea” and rejecting anything that doesn’t align.)
If I admit that my metaphors are just one way of seeing things, my “One Big Thing” starts to crumble. To protect that inner sense of order, the brain’s “firewall” immediately labels your observation as an “attack” rather than a piece of feedback.

4. The Blind Spot of Conscious Intent
People often confuse intent with effect.
They think, “I don’t intend to be a prisoner of language, so I’m not.” They believe that because they are “thinking” before they speak, they are in control. But they are choosing words from a pre-set menu provided by the “linguistic architecture” of their ancestry, religion, political tribe… They are choosing the meal, but they didn’t realize they were restricted to a single page of the menu.