Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I worked as a software engineering tech lead on the "Ground Truth" team at GM Cruise for a short period in 2018.

During orientation, the founder (Kyle Vogt) told us all that the cars don't need to be better than a human driver, they only need to be "not worse" than a human driver. In fact, when questioned about it, he said that the sooner we could get robot cars on the streets of SF the sooner the software could improve.

The idea was that the more dangerous, "not worse than a human" era would be a necessary sacrifice of safety (compared to waiting forever until we had 100% perfect driving robot cars) -- so we could fast-forward to major improvements in the robot cars' capabilities.

That (IMO) cavalier attitude, a lack of rigor in the way software for the vehicles was being developed, and the fact that we had constant meetings about "diversity and inclusion" rather than on robustness, safety, ethics and quality pushed me to resign abruptly and go my own way.

Very disappointing experience - as I had hoped to see "The Future" before joining GM Cruise and even a fat paycheck and RSUs weren't enough to sooth my scruples.



Not surprising that Cruise seems to cause an order of magnitude more issues than Waymo. It starts from the top down.


The classic "some of you might dir but that's a risk I'm willong to take?"




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: