The closer cars get to driving themselves, the we’ll become behind the wheel. It’s hardly a surprising thought, but it’s been supported once again by a new study from the and the Massachusetts Institute of Technology’s AgeLab.

For this experiment, 20 Massachusetts drivers were given one of two cars with differing assisted-driving technologies to use for a month. Ten were provided with a Range Rover Evoque equipped with adaptive cruise control, while the other 10 were supplied with a Volvo S90 outfitted not only with adaptive criuse control but also Volvo’s Pilot Assist system. Pilot Assist allows the car to steer itself within its lane as well as to adjust the vehicle speed for a safe following distance, making the S90 a Level 2 autonomous vehicle. The Range Rover, conversely, is classified as Level 1.

It probably won’t surprise you to learn that after the month was over, all participants were more likely to exhibit habits of “disengagement” compared with their driving of a vehicle without assistance technology. Disengagement is what the IIHS calls behaviors like removing both hands from the wheel or diverting attention from the road to use cellphones and adjust vehicle controls. The more alarming bit is how bad it got for the Volvo group, as IIHS senior research scientist Ian Reagan explains (important parts bolded by yours truly):

“Drivers were to show signs of disengagement after a month of using [Volvo’s] Pilot Assist compared with the beginning of the study,” Reagan says. “Compared with driving manually, they were to take both hands off the wheel after they’d gotten used to how the lane centering worked.”

This is an inevitability. The road to full autonomy is very long, and until carmakers get to the end of it we’re going to keep seeing progressively more capable technologies that automate certain aspects of driving but never the whole act. Of course, drivers must always be at the ready to assume control in case human intervention is suddenly needed. But that’s fundamentally at odds with the reasons people want self-driving cars in the first place, which is to do everything but drive.

Communication between assisted driving systems and drivers could also use some improvement. A few months ago, I drove a Hyundai Sonata Hybrid equipped with the Highway Driving Assist feature. I took my hands off the wheel on a few occasions to determine how long the car would do its thing before demanding my full attention. Sometimes it’d take 10 seconds; other times, more than half a minute. I was never able to determine the reason for the inconsistency.

However, there may be a silver lining here. Remember the group driving the Range Rover Evoque that had only adaptive cruise control? Turns out that while those subjects did pick up some ill-advised habits, the particular behaviors weren’t as careless, and those drivers were no more likely to stop holding the wheel at the end of the month:

Drivers of the Evoque, who used ACC often, were more likely to look at or pick up a cell phone while using the assistance technology than when driving manually, and that tendency increased substantially as they grew familiar with ACC. On the other hand, increased familiarity did not result in more frequent texting or other kinds of cell phone manipulation known to increase crash risk (see “,” Dec. 10, 2015). Unlike drivers using Pilot Assist, drivers using ACC in both the Evoque and the S90 weren’t any more likely to remove both hands from the wheel than when driving manually.

Perhaps there’s a sweet spot — a “just right” degree of assisted driving tech that makes road travel safer but still requires enough engagement that drivers can’t ever completely nod off. It’s a nice thought, but the way modern cars are marketed, and the expectations buyers have of them, suggests that this problem is going to keep getting worse before it gets better.