This week has had a few interesting news which, perhaps, suggest that the autonomous car fantasy may be starting to slowly come appart.

First, Google spun off its self-driving car tech as a new company, itself planning instead to

work alongside other automotive companies to develop cars that include “some self-driving features”.

That sounds really strange. Why scale back on expectations in this particular way?

Then, one of Uber’s autonomous cars was caught on camera running a red light.

Obviously, only the guy sitting in the car at that point knows what happened, but my guess is this: this car was running with prototype ‘‘level 3 autonomy’’, and the driver just got distracted. Level 3 autonomy means:

Within known, limited environments (such as freeways), the driver can safely turn their attention away from driving tasks.

The catch is that it should be ready to take over when things get out of hand. Personally, I think this is more dangerous than a plain old car without any electronics in it. Because suppose that what you do is taking a nap. Or texting on your smartphone. So you are completely unaware of traffic when suddenly the car calls for your attention. How long will it take you to intervene? And what if the car just goofs and doesn’t call your attention? How long until you notice?

So maybe these two pieces of news mean nothing. But maybe at google they’ve understood that this driverless car isn’t going anywhere (no pun intended), and maybe Uber is about to learn that a halfway solution can be worse than no solution.