Arrow of time and its reversal on the IBM quantum computer

Essentially it seems like they are exploring ways in which for certain computations a solution can be found before the computation has actually finished, approaching something like O(1) performance.

I think Scott Aaronson already weighed in on this:

https://www.scottaaronson.com/blog/?p=4147

I think a non clickbaity version of this article would look something like the following:

Note that "time reversal" is a common accurate technical term. The problem is that you get one or two clases in an advanced QM class explaining what it mean and what it doesn't mean. Using it in a title in a non technical publication is misleading (or clickbaity).

It's easy to explain what they did with an unrelated experiment:

Let's imagine that you make an experiment where you have a few ball laying on a billiard table. You hit one, it colide with the other and you wait tome time "t". Let's assume that the walls are magic and the ball bounce perfectly without loosing energy.

Now with some very accurate device you measure the position, speed and spin off all the balls, and then with some combination of magic and very precise actuators you create an scenario where all the ball have that positions, with the speed in the exact oposite direction, and spinning with the oposite spin.

If you wait again a time "t" you will see that everything goes in the other direction and the ball return to the initial position, like a movie shown backwards. So it's like "reversing the time", but you don't reverse the time in the "clock in the wall", you only make the system apparently move backwards.

With a quantum system with particles or pseudoparticles it is easier to have perfect collisions and perfect walls that don't loose energy, but you can't measure the position and speed at the same time, so you can't recreate the scene like in a classical system. Also, the balls may be entangled and have other nasty quantum stuff.

The trick they did is to use a quantum system to simulate a quantum system. The operation to reverse the direction is related to complex conjugation. IIRC they simulate the real and imaginary part separately to make it posible to reverse the imaginary part and simulate the complex conjugation. The good part of the complex conjugation is that it work without measuring the position and speed at the same time that is impossible in a quantum system.

So they star an experiment, wait "t" time, take the complex conjugate, wait "t" time and (most of the time) they return to the inicial configuration.

The second part where things move backwards is call the "time reversal". They are not making the wall clock go backward, only the part of the system move backward.