We Have More Evidence Than Ever, and Somehow Agree on Less!!!

A few months ago, I watched a video of an argument online.

It was maybe forty seconds long.

Someone had recorded it from across the street. There were two people shouting, a few others standing around, and then one person suddenly pushed the other.

I watched it once.

My conclusion seemed obvious.

Then I opened the comments.

Someone had posted another video.

Same argument. Different angle.

That changed things.

Then someone else wrote that something had happened before the first recording started. Another person claimed to know one of the people involved. Someone posted a screenshot. Someone said the screenshot was old. Another account uploaded a longer version of the video.

Ten minutes earlier, I had watched something with my own eyes and thought I knew exactly what had happened.

Now I wasn’t even sure what the argument was about.

And that’s a strange problem for an age that records almost everything.

We thought more evidence would make things clearer

For most of human history, enormous events could happen with surprisingly little documentation.

People relied on witnesses, newspapers, official records, photographs if they existed, and eventually television footage.

Today, someone can trip over a chair in a café and there’s a reasonable chance three camera angles will exist before they’ve stood back up.

We carry high-resolution cameras everywhere.

Cars have cameras.

Shops have cameras.

Doorbells have cameras.

Streets have cameras.

Satellites photograph the planet.

Messages can be screenshotted. Speeches can be replayed. Public statements can be searched years later.

You would think this would make disagreement harder.

Surely once everyone can see the evidence, reality wins.

Instead, something almost opposite seems to have happened.

We have become extraordinarily good at documenting reality and increasingly bad at agreeing on what the documentation means.

A video feels like the truth, but a video has edges

Every recording begins somewhere.

And every recording ends somewhere.

That’s obvious, but we forget it surprisingly easily.

A 20-second clip might perfectly show what happened during those 20 seconds while telling you almost nothing about the five minutes before them.

Move the camera slightly left and another person appears.

Start recording thirty seconds earlier and the person who looked like the aggressor might suddenly look like the one responding.

Add another hour of context and perhaps the entire interpretation changes again.

None of the videos necessarily has to be fake.

That’s what makes this difficult.

Something can be completely authentic and still give you the wrong understanding.

We spent years learning to ask, “Is this real?”

Maybe the harder question now is:

“Real… but how much of it?”

Then evidence became content

Something else happened when information moved onto feeds.

Evidence stopped arriving alone.

It arrived with a caption.

“THIS IS INSANE.”

“Nobody is talking about this.”

“Watch what happens at 0:17.”

“This changes everything.”

Before you’ve watched the video, somebody has already suggested what you’re supposed to notice.

Then you scroll into the comments and hundreds of strangers offer interpretations.

Then an algorithm notices you stayed.

So it gives you another video about the same event.

Then another.

Very quickly, it can feel as though you’ve independently investigated something.

But perhaps you’ve actually seen the same interpretation six times.

Repetition has a strange relationship with truth.

Something doesn’t become more accurate because we encounter it repeatedly.

It just starts feeling familiar.

And familiar things are dangerously easy to mistake for obvious things.

The same evidence can now support completely different realities

Imagine two people opening their phones after the same major event.

One person’s feed contains videos showing angry crowds.

The other’s contains peaceful crowds.

One sees clips of police violence.

The other sees clips of violence against police.

One sees experts saying the situation is improving.

The other sees experts saying it’s collapsing.

Here’s the uncomfortable part:

All of those clips might be real.

Neither person necessarily needs to have been “fed fake news.”

They may simply have received different pieces of reality.

Then they meet.

Both have evidence.

Both have videos.

Both have screenshots.

Both can say, “Look. It’s right there.”

And both genuinely wonder how the other person can possibly be so ignorant.

That’s a much harder information problem than simply separating truth from lies.

We used to argue about facts. Now we can argue with facts.

There was a hopeful idea in the early internet era that access to information would make society more informed.

Put knowledge in everyone’s hands.

Let people check claims themselves.

Make information impossible to control.

A lot of that happened.

And it was enormously valuable.

But we underestimated something.

Giving everyone access to evidence also gives everyone access to an almost unlimited supply of evidence supporting whatever they already suspect.

There are billions of videos, articles, charts, posts, studies, interviews and screenshots online.

If you begin with a conclusion and search long enough, you’ll probably find something that looks like proof.

Maybe the information age didn’t eliminate confirmation bias.

Maybe it gave confirmation bias a search engine.

And now there’s another complication

For a long time, seeing was believing.

Then photographs became easy to manipulate.

Now video and audio can be generated too.

AI doesn’t need to make the entire internet fake to create a problem.

It only needs to make fake things plausible enough that real things become deniable.

That’s a subtle but important difference.

Imagine genuine footage appears showing someone doing something terrible.

They don’t necessarily need to prove the footage is fake.

They only need to say:

“That’s AI.”

If enough people already want to believe them, uncertainty does the rest.

The danger isn’t simply that we’ll believe things that never happened.

It’s that we may stop believing things that did.

Maybe evidence was never the final problem

We assumed disagreement existed because people didn’t have access to the same information.

Sometimes that’s true.

But people don’t experience information like machines receiving data.

We bring history.

Identity.

Trust.

Fear.

Politics.

Experience.

Communities.

Expectations.

Two people can look at exactly the same thing and notice completely different parts of it.

So perhaps the biggest information problem of our time isn’t that we don’t have enough evidence.

We are drowning in it.

The problem is deciding what deserves context, what deserves trust, and what might be missing from the frame.

That’s slower than scrolling.

Less satisfying than certainty.

And much harder than choosing a side.

But maybe the most intelligent sentence on the internet right now is also one of the least impressive:

“I don’t know enough yet.”

In an age where everyone can produce evidence instantly, being willing to wait before deciding what it means might become a surprisingly valuable skill.