The first is that we're more likely to get motion sickness when our inner ear is telling us something different from our eyes (or, as Sivak puts it, when there's a conflict between our visual and vestibular inputs). That means, for instance, that you may feel queasy when you're reading a book in the back seat of a car, because your eyes tell you you're not moving, but your inner ear insists that you are. This is also why your stomach may turn during a high-speed chase at the IMAX — your eyes think you are moving, but your inner sense of balance disagrees.
In a driverless car, we'll all be passengers. And to the extent that this will enable many of us to read books in the car, or work on our laptops, or troll the Internet on our smartphones, we'll be more prone to motion sickness than we are when we're at the wheel following the road.
We're also more likely to experience motion sickness when we can't anticipate the direction of motion (when a vehicle suddenly turns right or changes lanes), and when we don't control the direction of motion ourselves. Which is pretty much what will happen in driverless cars.
All of this means, by Sivak and Schoettle's calculation, that somewhere between 6-10 percent of American adults riding in fully autonomous cars will experience some kind of motion sickness regularly. Their calculations here are based on estimates about what we'll be doing while we're riding in these cars, since our multi-tasking will have a lot to do with when we get motion sickness.
In short: If you're looking out the window as the car drives you around, you may well be safe (this is why drivers seldom get motion sickness). But if you're sitting there reading The Washington Post on your 10th generation Kindle? Then you may have a problem that technology cannot solve.