Not paywalled for me, so I’ll share the relevant part:

Gravity is the weakest of our four fundamental forces—the other three being electromagnetism and the strong and weak nuclear forces. Gravity is unusual among forces, not only because of its weakness, but also because it’s always attractive between all objects. In other words it acts as a universal glue. That’s exactly what keeps us stuck to Earth, keeps Earth in orbit around the sun, and keeps the sun in its own orbit around the center of our Galaxy, and so on. But gravity’s quantum nature could actually be used to make gravity repulsive. This is my latest work with Marletto and another colleague, Pablo Saldanha, in which we designed an antigravity machine.

The machine works similarly to the BMV experiment. Imagine that one of the two gravitating masses (the “source”)—or a beach ball, in terms of our layman’s experiment—is in a superposition of states, while the other (the “probe”) is localized in one place. In the part of the superposition where the source is closer to the probe, the gravitational attraction is stronger. Meanwhile, in the other part where the source is farther away from the probe, the gravitational attraction is weaker. In both branches, the force of gravity is still attractive, so how do we make this into repulsion?

For starters, every quantum experiment has three parts:

1. Prepare a superposition.
2. Let it evolve in time.
3. Finally, measure in another superposition.

It’s this last part that gives us the repulsion, but it only does so for one of the outcomes of the final measurement. So what matters is the post-selection; we have anti-gravity only if the right outcome is observed. On average, if both outcomes are included, gravity is always attractive, just as it is in the classical world. So, we need to discard one of the outcomes of the final measurement, which is what gives us repulsion.

While some might argue that anything can happen if we post-select—or observe the most favorable outcome—that isn’t necessarily the case. Indeed, discarding “bad” outcomes (such as those where the particles attract) leads us to observe what we want (in this case, repulsion). However, classically, this isn’t possible, no matter how much we post-select. If our experiment is confirmed, it would therefore show that gravity can act from two different points on the source at the same time. In other words, only if gravity is quantum could we have an antigravity machine.

How realistic are the BMV and antigravity experiments to perform? Pretty difficult. Luckily, a number of world-leading quantum groups are racing to implement these experiments, and I’m optimistic that we will have conclusive results in the early 2030s. Even more excitingly, I am collaborating with Marletto and my Italian colleagues Marco Genovese, Fabrizio Piacentini, and Ettore Bernardi, who are wizards in the lab, to try to get there first. We are on the cusp of solving one of the biggest mysteries of physics, and, as a bonus, might be able to develop technology that even renowned science-fiction writer Arthur C. Clarke couldn’t have imagined.```