The first description of a quantum entropy feedback amplifier-actuator (patent pending):

It might be possible to manipulate the normal laws of probability.
According to the Many Worlds interpretation of quantum physics, our timestream keeps splitting into multitudes of possible outcomes. Every possible subatomic quantum event, no matter how small, causes a new universe to 'pop' into existence, complete with memories of its own past. In a small fraction of new universes, everything will have 'gone right' by pure chance, meaning any unlikely but not impossible desired outcome will have come to pass somewhere. The near-miracle that we wanted to happen actually did happen there. A handful of coins always comes up heads. Moving randomly, atoms just happen to rearrange themselves into any desired object.
It might be possible to amplify these extremely rare possibilities, so they would outnumber all other undesirable timestreams.
This would violate the laws of entropy as we know them. It might be necessary to first increase entropy levels elsewhere, perhaps inside stars or empty space, or by slightly speeding up the expansion of the universe.

How it would work:
Observers in many universes would experience many random outcomes.
The most successful ones would then take steps to 'multiply' the universe they inhabit. The less successful observers would do nothing, or try again until they were successful.
A selected universe could be multiplied by rapidly increasing its number of possible future states. This could be done by 'releasing' the disorder in a formerly highly symmetrical system; for example by unleashing a large amount of energy in a chaotic way.

One way to test if such a device is possible would be to detonate a large number of nuclear devices (as big as possible), each controlled by a random ten-second timer.
After a large number of detonations, timing deviations might become noticeable. On average, most detonations might occur slightly earlier than expected during the ten second interval.

If it works, it would tell us something about our multiverse, and about reality as a whole.
According to most standard theories, this method would never work, but then the experiment has never been conducted.
For example, there might always be an 'opposite' possible universe to counter every probability we try to amplify.

Perverse paths:
This feedback process would have hidden dangers. Even if an outcome looks perfect, it would still be fundamentally random.
Even after a thorough inspection, there would be an overwhelming probability that as many hidden faults as possible remain in the seemingly desired outcome.
The moment the feedback process stops, the contrived situation might fail at the earliest possible moment.

   Developing a quantum entropy feedback device

According to some versions of the many-worlds interpretation of quantum mechanics,
when a system collapses from an ordered to a more chaotic state (like during an explosion), it causes the universe to 'split up' and multiply into many different versions of itself, in which each possible outcome is realized.
During this early period in the history of our cosmos, only fourteen billion years after the Big Bang, the universe is far more likely to split up than to recombine.
There are many more ways for our still-ordered universe to become more chaotic than more ordered, which also causes our 'arrow of time'.

In the following thought experiment, a highly ordered universe contains only one mind (or relatively few), waiting for an 'entropy bomb' to go off or not. It might be triggered by some random event or number generator.
The subsequent more chaotic universes in which the bomb did go off contain many nearly-identical minds, each witnessing a slightly different version of the explosion.
According to the anthropic principle, a mind is more likely to exist in one of the chaotic universes after an explosion, than in one of the universes where the bomb did not yet go off.
Wherever there are more possibilities, there are more minds.

If true, this could become a powerful new tool beyond anthropic computing and quantum immortality.
Suppose you want a certain event to happen, say a coin to come up tails multiple times. Simply decide in advance to detonate a 'quantum bomb' only if the desired outcome occurs, and then follow through.
The 'bomb' would have to be a high-energy, entropy-unleashing event. Thoughts or perceptions alone probably aren't powerful enough to manipulate probability. The brain uses just as much energy when it's thinking as when it's resting.
Many more universes will have 'split off', containing more observers, in which the chosen event happened to take place, creating the illusion that luck itself can be amplified.

According to standard physics, the above procedure would never work. The math is rather complicated.
Maybe the total number of parallel universes can't change. Infinite to begin with, they can only split into sub-sets.
Anthropic entropy may 'cancel out', because in some other universe, some other copy of yourself will always do the exact opposite thing.
If a quantum bomb was possible, biological evolution would have already exploited it. But what if the experts are wrong? No one fully understands quantum physics.

A proposed experiment:
------------------------
Previous quantum experiments haven't detected any statistical anomalies, but perhaps their tools just weren't powerful enough. The effect may only be noticeable in a very large energy release. A few hundred powerful hydrogen bombs could be simultaneously detonated on the far side of the moon, where their radiation wouldn't threaten Earth.
Each bomb would be programmed to have a 50% chance of detonating during a ten second interval, controlled by a Geiger counter. Those bombs that don't go off are later defused and destroyed. This experiment would link a complex macroscopic state to a single quantum transition.
It would be the largest possible quantum amplification event, except for supernovas or the Big Bang itself.
If the many-worlds interpretation is correct, and if the anthropic principle can be used to manipulate the apparent outcome of quantum splitting, more bombs would detonate than not. Perhaps all of them would go off.
According to all standard theories, only about 50% should explode.
If such an effect exists at all, it may be very small and hard to detect.

The next step would be to test if the detonations could be used to manipulate the apparent outcome of a coin toss or a random number generator. The bombs would only go off if a desired event had taken place. Perhaps most observers would then find themselves in realities in which nature appears to conspire to set off the quantum explosions. More likely, many trials would be needed to detect a statistical deviation. However, if the phenomenon is real, it could be self-amplifying.
In that case, a single quantum bomb could manipulate the outcome of many artificially linked events.

---

Probably the hard SF novel ever written: Infinite Thunder by Jack Arcalon.
Enigmatic source of the Anonymous meme (really).
Buy the book
Read the chapters