"Are you familiar with the double slit experiment? Individual particles are fired at a piece of metal with two thin slits in it. Let's use photons as an example. Behind the slits, a detector records the location each photon hits.
Over time, an interference pattern develops. This means that the photons are acting like waves - interfering with themselves, since only one photon has been sent through at a time.
Now if a detector is placed in front of one of the slits to see if the photon is going through that slit, the interference pattern disappears. By looking to see if the photon is going through the slit, you apparently force it to choose. This is the central mystery of quantum mechanics.
If you put a detector up, you do not get to choose which slit the photon goes through. If you do not put a detector up you do not control where on the observation strip the photon hits.
Here is where we deviate from universally accepted physics. We created a machine that can influence this 'choice.' The Machine can force which slit the particle goes through.
How to make this useful, though? It's only noticeable in large sample sizes, and there are plenty easier ways to direct these particles. Yes, we have something that would electrify all physics. We can manipulate quantum mechanics on a supposedly impossible level.
The quantum effects, however, are swamped at larger scales. The small statistical effect becomes meaningless. We tried improving the magnitude of the effect, and succeeded.
The improvements came at the expense of the reliability of the Machine. It malfunctioned in one way after the other. Eventually we disassembled it. We tested every part, then tested every subsystem, then tested all the subsystems together without actually activating the machine.
I'm in charge of these tests. It's not really my speciality, but that's seniority for you. This is important: the Machine worked fine in these tests, but when the Machine was actually used reliable parts would fail, and strange edge case interactions would crop up. Every component malfunctioned at a vastly increased rate.
I think I know why. The Machine malfunctioned because it was more likely that the Machine malfunction than all those photons go through the chosen slit."
A rustle of paper comes over the recording. A few words from someone too far from the microphone. Dr. Tanaka continues. "I see what you are thinking. You are thinking this is my delusion. Why I have come to you. If you check the psychoanalyses that I provided, you will find this isn't my problem."
The doctor's voice came again, louder now. "I'm not making any judgements. I hope you can trust me that much. Please continue at your own pace."
"What does this mean? Our best guess is that the Everett many worlds interpretation is correct. Or at least closer to correct. Each time a photon goes through the slits the universe splits. In one world the photon goes through the right slit, in the other it goes through the left slit.
Somehow, the Machine can select which universe 'we' are in. So when we set the machine to 'left,' we could be in a world in which the Machine functions correctly, and see the result 'left,' or we could be in a world in which the Machine malfunctions, and see some random result.
Let's say the chance the Machine malfunctions is 1%. If we have ten photons we want to go left, then the universe will split ten times. There will be 1024 worlds with different results. In only one of those 1024 worlds does every photon go left.
So what? Why don't we always find ourselves in that universe? Because there are ten universes where the machine malfunctioned. So in ten out of eleven worlds the machine malfunctioned.
My colleagues think that the Machine entangles its state with that of the photon. This explanation could be extended to the Everett interpretation. The Machine in effect moves the split point earlier - to when the Machine is programmed to select right or left. In one world we select right. In one world we select left.
I do not believe this is true. You see, I have looked at the data from when the photons were flowing. From when the machine was on, but there was no test being run. A pattern that should be random. To be sure, some of the data was random. The early parts. The later parts, however...
Tell me: do you know the sensation of insanity? It sounds strange, but I have felt it. Imagine your mind rotating around a spindle that has jumped its socket. You can think and plan amazingly, but at the root something has slipped.
I'm not describing it well. I looked at that random noise and felt that sensation. I investigated and found a pattern. When I was young I made a childish cipher. On a whim I tried decoding the first of the data using this cipher. It said 'Hllo.'
I tried on the rest, but it didn't work. I tried dozens of different things, until eventually I tried a symmetric key passphrase I had memorized, but never used. Interpreting the output as characters didn't work, but as numbers I found a message.
Once again I looked carefully. The numbers varied from 0 to 1457. There were many more low numbers than high numbers. It took a while, but I cracked this puzzle too.
How would you convey a message efficiently?"
He paused for a moment, as if expecting an answer. None came, and he continued impatiently. "The numbers corresponded to the most commonly used words in English, in order of frequency. This, finally, was the message in the static. Still difficult to understand. Frustratingly short and condensed, but it was a message from me to me."
"What did the message say?"
"Oh, various things. It was very short. Even running the machine continuously during off hours yielded only a few bits at night. Most importantly, it contained hints. Very short hints that inspired ideas to improve the Machine."
"Do these ideas work?"
"I haven't been able to introduce most of them yet."
"These ideas are things you could have come up with on your own, right?"
"Yes. The message gives just the bare minimum hint necessary to lead me to them."
"Dr. Tanaka: you are clearly a very intelligent man. I'm sure you've heard of the conspiracy theorists who find secret codes in Shakespeare, or the Bible. They use methods similar to the what you described - trying so many things in so many combinations that they eventually find something that appears meaningful.
It is common for schizophrenics - not that I'm saying you are schizophrenic, just giving some examples - to see hidden messages to themselves in the media. Humans are very good at seeing patterns, and sometimes that strength becomes a weakness.
You ascribe patterns to nothing, see hints about things you are already thinking about. It never gives you a truly novel idea, just those you could reach on your own. What does that tell you?"
"Yes, I am aware. Why do you think I have come to you?"
"That was a very good idea, I'm glad you had the bravery and mental fortitude to seek professional help."
"Well thank you. I will take that in the spirit in which it was given. However that was not the totality of my meaning. This is a test too. Of me, and of the message. Look out the window. Right above that little mountain peak."
"What am I looking for?"
"A meteor. If the message is correct there will be a large meteor there in six minutes."
"And if there isn't? Will you agree that this message is a delusion, and accept whatever treatment you need?"
"As I was saying, I realized the simplest explanation was that the Machine didn't select which world we were in, but that it removed all worlds that didn't fit its criteria. Naturally, we could never find ourselves in a world line that didn't exist, so to us it appeared that the Machine forced the photon to go left.
Whether those worlds never existed, or existed and were destroyed by the Machine is not discernible. Practically, it makes no difference, though perhaps it does philosophically.
The Machine removed the world-lines. Do you know what that meant? It meant that you could order the Machine to remove any world where it rained on Sunday. Even better, you could do this in the future, and it would apply to the past. The Machine would remove any world-line where it rained on Sunday from the moment it was turned on to the moment it was turned off, regardless of when you gave it the actual order.
So my future self simply told it to remove any world-line where the first bit wasn't 1, and then remove world-lines where the second bit wasn't 0, and so on. Do you understand the implication? If I receive a message, then some time later send it back to myself, where did the message come from?
This question, logical as it may seem, indicates a flawed understanding of reality. The message came from nowhere. It is a possible message, so it might be. The important question is what makes a message more or less likely. The world-line that encourages its own creation and monopoly will be the most likely survivor. The message will be the one that most encourages its own transmission.
Something is going to whisper back in time and bootstrap itself into existence. Something is going to take over the world, all the worlds. I happened to have a huge advantage. The first to move wins, and I was in the perfect spot.
I am competing with everything - every possible monster. Every alien, incomprehensible future threat. Every superhuman intelligence, every eldritch god. Anything that could possibly exist.
Here is their chance to exist, with only me to stop them. The only way to do so is to become something similar. To be better at monopolizing the world-lines than they are."
The psychiatrist sounded somewhat confused. "I must say I am glad you have opened up to me. Our previous sessions haven't been nearly as productive. You are doing the right thing."
"Am I? I once promised myself that I would always be a hero. I might have to do terrible things to save the world. The hero is not the one who does such things, even for that cause. The hero is the one who refuses to do evil, no matter what.
Villains can do the wrong things for the right reasons. I might have to break that promise to myself, but I will not forget my shame."
Careful concern tinged the psychiatrist's next words. "What terrible things?"
"I was willing to open up to you because one way or another I would have my answer today."
"Yes, but it is important that you explain what you mean by having to do terrible things."
"Not now, look at the sky. Five seconds."
A few seconds passed on the recording with only light static. Then a sudden intake of breath. "It's there! What! Did you set that up somehow?"
"No. Would you agree that it split into three pieces? Three trails?"
There was no answer but a heavy crashing noise. There was an exclamation and several thuds. The sound of heavy breathing and an indistinct shout. Then Tanaka's voice:
"I am so sorry, but you know too much."
There was a gagging noise, but no reply.
"Don't worry, you can be confident your death is part of the plan that saves the world. Don't you understand? My exalted future self has granted me foreknowledge. I know that I succeed, for I am already victorious."