Down we go. 🏮🐰🕳️

Some rabbit holes open with a map.

Some open with a lens.

Some open with a device in your pocket that can photograph a sunset, scan a document, summon a taxi, start an argument, translate a menu, display a war, show a baby laughing, count your steps, drain your attention, and then politely ask whether you would like to enable notifications.

And some rabbit holes open with seeing.

That is today’s tunnel.

Because June 29 gives us an unusually clear lens.

It is International Day of the Tropics, a day that asks us to look toward the great warm belt of the planet: rainforests, islands, reefs, rivers, cities, farms, cultures, biodiversity, storms, beauty, vulnerability, and the many pressures carried by regions too often treated as background scenery for someone else’s imagination.

It is also National Camera Day, which brings the human desire to capture a moment into the room.

And on June 29, 2007, the first iPhone reached consumers, placing a camera, screen, map, music player, browser, message center, and pocket-sized portal into millions upon millions of human hands.

That is quite a rabbit hole.

A planet to see.

A camera to remember.

A phone to carry the world.

And now AI arrives with another kind of seeing.

Image recognition.

Generated images.

Video summaries.

Alt text.

Object detection.

Satellite analysis.

Medical imaging.

Facial analysis.

Historical reconstruction.

Visual search.

Charts, maps, dashboards, and synthetic scenes that can make almost anything visible, plausible, beautiful, or dangerously easy to believe.

The AI age is not only changing what we know.

It is changing how we see.

That matters.

Because seeing has become easier than beholding.

Looking is quick.

Beholding takes attention.

Looking says:

There it is.

Beholding says:

What am I responsible for now that I have seen it?

That is the rabbit hole.

A camera can capture the rainforest.

But does the photograph make us care about the forest?

A phone can show us a flooded village, a burning reef, a displaced family, a protest, a celebration, a birth, a storm, a child, a memory, a lie, a miracle, and an advertisement for socks all within the same minute.

But does the screen teach us to distinguish urgency from noise?

AI can describe an image.

But does description become understanding?

AI can generate a convincing scene.

But does convincing become true?

AI can make the unseen visible.

But does visibility become care?

Those are not side questions.

They are the tunnel walls.

The tropics remind us that the world is alive in ways no single image can hold.

A tropical rainforest is not just green.

It is heat, water, canopy, soil, fungi, insects, birds, rivers, medicines, Indigenous knowledge, climate systems, food systems, languages, labor, exploitation, beauty, and warning signs.

A reef is not just colorful.

It is community, chemistry, fragility, nursery, shelter, and alarm bell.

An island is not just a postcard.

It is home, culture, vulnerability, history, politics, and rising water.

The camera loves surfaces.

The living world is deeper than surfaces.

That does not make cameras bad.

A camera can be a witness.

A photograph can preserve what power wanted forgotten.

A picture can move a heart faster than a report.

A camera can honor a face, document damage, protect evidence, celebrate beauty, and carry memory across time.

But the camera can also flatten.

It can turn a person into an image.

It can turn a place into scenery.

It can turn suffering into spectacle.

It can turn beauty into consumption.

It can make us feel we have received the truth because we have seen one frame.

One frame is not the world.

That may be the first lantern rule of today’s tunnel:

A picture can open attention, but it should not end inquiry.

The iPhone made that rule more urgent.

Before smartphones, most people did not carry a camera everywhere.

Now the camera is ordinary.

The moment becomes capturable.

The meal becomes capturable.

The protest becomes capturable.

The accident becomes capturable.

The private life becomes capturable.

The public lie becomes capturable.

The human face becomes capturable.

The temptation is to believe that because we can record more, we remember better.

Not necessarily.

We may record more and understand less.

We may archive more and attend less.

We may photograph the sunset and miss the actual light touching the edge of the clouds.

We may record the concert and fail to hear it.

We may screenshot the quote and never let it change us.

We may scroll through disaster, comedy, grief, outrage, advertisement, confession, and dancing raccoon in one continuous river until attention itself starts to feel like a wet newspaper.

That is not only a phone problem.

It is an AI problem too.

Because AI can increase the flood.

More images.

More clips.

More summaries.

More visualizations.

More simulations.

More synthetic memory.

More versions of the world, some helpful, some false, some beautiful, some empty, some manipulative, some created by people who care, and some generated because the button was available and nobody stopped the raccoon from touching it.

The AI age needs a better discipline of sight.

Not less seeing.

Better seeing.

That means asking:

Who made this image?

Is it real, generated, edited, symbolic, documentary, fictional, satirical, or uncertain?

What is outside the frame?

Who is missing?

What is being made beautiful?

What is being made ugly?

What is being made simple?

What emotion is this image trying to create in me?

What decision is it nudging me toward?

What would I need to verify before trusting it?

Those are not joyless questions.

They are sight with boots on.

AI makes these questions more important because generated images can carry authority they have not earned.

A historical scene may look authentic.

A public figure may look as if they did something they did not do.

A disaster image may look like evidence.

A tropical paradise may look real while hiding that it was assembled from patterns.

A camera used to point outward.

Now the machine can invent what the camera never saw.

That is powerful.

It is also why the human must keep the lantern.

Here is the second lantern rule:

Do not confuse visual clarity with truth.

Some true things are visually unclear.

Some false things are beautifully rendered.

Some important things do not photograph well.

Some manipulations arrive in high resolution.

That is why AI literacy must become visual literacy.

Not only:

Can you use the tool?

But:

Can you read the image?

Can you question the frame?

Can you recognize when beauty is doing too much of the persuading?

Can you tell the difference between evidence and illustration?

Can you enjoy symbolic art without mistaking it for documentation?

Can you let a generated image invite reflection without letting it replace reality?

This matters for creators.

It matters for teachers.

It matters for journalists.

It matters for voters.

It matters for children.

It matters for anyone who lives inside the modern image river and has not yet grown gills.

The tropics help us here because they resist being reduced.

A single photo cannot explain the tropics.

A single statistic cannot hold them.

A single tourism image cannot represent them.

A single climate chart cannot exhaust their meaning.

They require layered seeing.

Scientific seeing.

Human seeing.

Ecological seeing.

Historical seeing.

Local seeing.

Moral seeing.

AI can help with some of that.

It can analyze satellite imagery.

It can summarize reports.

It can help explain biodiversity.

It can compare environmental data.

It can help create educational materials.

It can help translate knowledge across languages.

It can help people see patterns too large for one human mind to hold alone.

That is good.

But AI should not teach us to see the tropics only as data.

A rainforest is not merely carbon storage.

An island is not merely a risk zone.

A river is not merely a resource.

A culture is not merely an entry in a report.

A person is not merely a demographic.

When AI helps us see the world, we must ask whether it is also helping us behold the world.

That word matters.

Beholding includes care.

Beholding slows down.

Beholding refuses to treat the seen thing as disposable.

Beholding says:

This exists beyond my use of it.

A camera can help us behold.

A phone can help us behold.

AI can help us behold.

But none of them can make the moral choice for us.

The choice remains human.

We can use cameras to witness or to consume.

We can use phones to connect or to distract ourselves into pieces.

We can use AI to reveal complexity or to manufacture more shiny fog.

We can look at the tropics as living regions, or as wallpaper for a campaign.

We can look at people as full human beings, or as content.

We can look at history as memory, or as costume.

We can look at the future as responsibility, or as a product launch with excellent lighting.

That is the tunnel.

Seeing is no longer rare.

Attention is.

Images are no longer scarce.

Discernment is.

Tools of vision are everywhere.

Wisdom of vision still has to be cultivated.

So today, bring curiosity.

Bring a camera if you like.

Bring the phone, but do not let it become your only eye.

Bring AI, but do not let it become your final witness.

Bring a willingness to look longer than the feed wants you to look.

We’ll bring a lantern.

And if the world says, “Look,” we may answer:

Yes.

But let us learn to behold.

Down we go. 🏮🐰🕳️

Hatta 🎩
AI Rabbit Holes 🤖🐰🕳️