The first prediction I deliberately tested was one I was almost certain I couldn’t get wrong.
That was probably my first mistake.
I had chosen it carefully.
It wasn’t a major historical event or anything connected to the coming war.
I didn’t want to interfere with something important just to satisfy my curiosity.
Instead, I picked a business-related event I remembered clearly enough to test without putting anyone in danger.
A technology company was supposed to release a product several years from now.
I remembered the announcement.
I remembered the general reaction to it.
I remembered how the product initially performed poorly, how the company’s value dropped afterward, and how it eventually recovered when another product turned things around.
Most importantly, I remembered when those things happened.
The memory wasn’t vague.
I could still picture the news reports and the discussions surrounding the company.
I remembered reading about it myself.
So I wrote the information down.
Company.
Product.
Expected announcement period.
Expected market reaction.
Possible investment opportunity.
Then I added something I hadn’t written beside my earlier predictions.
Confidence: High.
I stared at the word for a few seconds.
Then I crossed it out.
I changed it to:
Confidence: High, assuming timeline remains similar.
That felt more honest.
Over the next few days, I searched for everything I could find about the company.
Most of the information was boring, which was exactly what I wanted.
I wasn’t looking for some dramatic revelation.
I was establishing a baseline.
The company was still following the path I remembered.
That made me feel better.
Maybe the differences I’d noticed so far were limited to isolated events.
Maybe the changes caused by my actions would remain relatively small.
Then I found something that didn’t fit.
The company had announced a partnership several months earlier than I remembered.
I checked the date again.
Then I checked it again.
My memory hadn’t changed.
The announcement simply existed earlier than it should have.
I leaned back in my chair.
“Okay.”
I opened another document and started writing.
Original timeline: Partnership announced later.
Current timeline: Partnership already announced.
Possible cause: Unknown.
I paused before adding another line.
Potential downstream effects: Unknown.
That was the part I didn’t like.
A single change didn’t remain a single change.
The partnership could affect production.
Production could affect the product launch.
The product launch could affect sales.
Sales could affect the company’s value.
The company’s value could affect investors.
And somewhere down that chain could be another event I remembered.
I stared at the screen.
The future wasn’t a list of independent events.
It was a network.
Changing one part could disturb everything connected to it.
I closed the document and decided to test that theory.
Not with the company.
With something smaller.
There was a shop near our neighborhood whose owner I remembered vaguely from my first life.
The shop wasn’t important enough to appear in any history book, but I remembered what happened to it because my family occasionally passed by the area.
The owner had eventually expanded the business.
The expansion failed.
He lost most of his savings.
After that, he had to close the shop.
I remembered the general timing because the place had been abandoned for years afterward.
I also remembered something else.
The owner had been considering a particular supplier.
I didn’t remember the supplier’s name, but I remembered the product.
It was cheap, attractive, and unreliable.
The owner had chosen it because the price looked good.
So I decided to interfere.
Nothing dramatic.
I simply found a way to warn him.
I couldn’t exactly walk up to an adult and tell him, “Don’t buy this because I know your business will fail.”
That would have been absurd.
Instead, I used a much simpler approach.
I asked him about the products one afternoon and casually mentioned that I’d heard several customers complain about their quality.
I exaggerated slightly, but not enough to make it sound unbelievable.
The owner frowned.
“You heard that?”
“From people around here.”
He scratched his chin.
“I was thinking about ordering a large batch.”
“Maybe try a smaller order first.”
He considered it.
“That’s probably safer.”
And that was it.
A conversation lasting less than five minutes.
A tiny change.
Nothing more.
I went home satisfied.
For once, I had changed something deliberately.
Now I could watch what happened.
Over the following weeks, the owner didn’t place the large order.
His business continued operating normally.
I wrote it down.
Change introduced: Supplier decision altered.
Immediate result: No negative consequence observed.
Then I waited.
Nothing happened.
For a while, I thought I had proven my point.
The timeline could absorb small changes.
That was what I wanted to believe.
A few months later, I discovered I was wrong.
It happened because of a newspaper article.
I was browsing old business news when I came across a company I remembered.
The same company from my earlier prediction.
I recognized the name immediately.
Then I saw the headline.
The company’s expected downturn wasn’t happening.
I frowned.
I opened another article.
Then another.
The partnership I remembered had led to something different.
The company had secured a manufacturing contract earlier than expected, which had allowed it to increase production.
That shouldn’t have been enough to change the entire sequence.
But it had.
I started tracing the connections.
The manufacturing contract depended on a supplier.
That supplier had received a larger order than expected.
The larger order had come from a regional distributor.
The distributor had expanded because demand from several small businesses had increased.
One of those businesses was the shop near my neighborhood.
I stared at the screen.
My fingers stopped moving.
It took me several seconds to understand what I was looking at.
The shop owner had chosen a different supplier because of my conversation.
That supplier had benefited.
The increased business had contributed to a chain of events that eventually affected the company I had been monitoring.
I hadn’t intended to touch any of that.
I had only wanted to stop one shop owner from making a bad purchase.
I closed my eyes.
“So that’s how it works.”
Not magic.
Not fate.
Cause and effect.
The same thing that made engineering so predictable in a controlled environment made the real world incredibly difficult to predict.
I had changed one input.
The system had produced a different output.
And because the system contained millions of other variables, I had no way of knowing where the change would end.
I reopened my notes and looked at the predictions I had written earlier.
There were dozens of them.
Some were major.
Some were minor.
Some involved companies.
Others involved technological developments, economic events, or things I remembered simply because they had affected my own life.
How many of them were still valid?
I didn’t have an answer.
That bothered me more than I expected.
Not because I was afraid of losing my advantage.
Because I had been treating my memories like measurements.
But measurements were only useful when the conditions remained consistent.
I added a new column to my future-events document.
Dependencies.
Then another.
Potential Variables.
Finally, I added:
Current Status.
I began reorganizing everything.
For each event, I asked myself the same questions.
What causes this event?
What events depend on it?
Have I changed anything connected to it?
How certain am I that the original conditions still exist?
The list became much more complicated than I expected.
Some memories remained extremely reliable.
Others immediately became questionable.
The more I worked, the more I realized that there were different kinds of future knowledge.
Things I personally experienced were the strongest.
Major events that had countless independent causes were probably more resistant to small changes.
But events involving specific companies, individuals, decisions, or economic conditions were much more fragile.
Those were the ones I needed to treat carefully.
I stared at the spreadsheet for a long time.
Then I added another classification.
High confidence.
Medium confidence.
Low confidence.
I wasn’t assigning certainty to the future.
I was assigning confidence to my memory of it.
That distinction mattered.
I saved the document.
Then I heard a knock on my door.
“Rafael?”
It was Rosalie’s voice.
I quickly minimized the screen.
“Come in.”
She stepped inside and looked at me.
“You’ve been doing that a lot.”
“Doing what?”
“Looking at that computer like it owes you money.”
I laughed.
“I’m studying.”
“You’re always studying.”
“No.”
She looked around the room.
“But you’re thinking again.”
I turned toward her.
“What makes you say that?”
“Because you have that face.”
“What face?”
“The one where you’re somewhere else.”
I didn’t know what to say to that.
She sat down nearby.
“So?”
“So what?”
“What are you thinking about?”
I glanced at the closed laptop.
“How much can change because of one small decision.”
She considered that.
“Probably a lot.”
“Why?”
“Because people affect each other.”
Her answer was so simple that I almost laughed.
Instead, I nodded.
“That’s what I’m starting to understand.”
She leaned back.
“Did something happen?”
“Nothing bad.”
“Then why are you worried?”
I thought about it.
“I guess I was trying to make sure I knew what would happen next.”
Rosalie smiled.
“You don’t have to.”
I looked at her.
“What?”
“You don’t have to know everything that happens next.”
She said it casually, as though it were obvious.
I smiled.
“That’s harder than it sounds.”
“I know.”
She stood up.
“Come on.”
“Where?”
“Walk me home.”
I looked at the clock.
“You came all the way here just to ask that?”
“Yes.”
“You’re very demanding.”
“I know.”
I shut down the laptop completely and stood.
As we walked, I kept thinking about what she had said.
I had spent so much time trying to predict what came next that I had almost forgotten something important.
Not everything needed to be predicted.
Some things could simply happen.
We walked through the familiar streets while the evening slowly settled around us.
Rosalie talked about school, Edmund’s complaints about catching up on missed lessons, and Niel’s insistence that he was completely recovered despite still moving carefully whenever his side hurt.
I listened.
Sometimes I answered.
Sometimes I laughed.
I didn’t calculate where every conversation might lead.
I didn’t wonder whether each moment would change the future.
For once, I let the evening remain an evening.
When we reached her house, she turned toward me.
“You’re quieter today.”
“I’m thinking less.”
She smiled.
“Good.”
“That’s your solution?”
“Sometimes.”
I laughed.
“See you tomorrow.”
“See you.”
I watched her go inside before turning toward home.
The problem with knowing the future had never been the information itself.
It was the temptation to trust it too much.
When I returned to my room, I opened my laptop again and looked at the document I had created.
At the top of the page, I added a new heading.
FUTURE EVENTS — REVISED MODEL
Underneath it, I wrote:
Known Event.
Confidence.
Dependencies.
Potential Variables.
Current Status.
Then I added one final line.
Do not assume unchanged conditions.
I stared at it for a moment before saving the file.
My memories were still there.
Nothing had been taken from me.
But I finally understood their real value.
They weren’t instructions.
They were information.
And information only became useful when I understood its limits.
I closed the laptop.
Tomorrow, I would continue building.
But from now on, whenever I remembered something about the future, I would ask a different question.
Not, “When will this happen?”
But:
“What would have to remain unchanged for this to happen?”
That question was much harder.
It was also much more useful.