GM indefinitely halts work on driverless car at Cruise unit | Hacker News Reader

GM indefinitely halts work on driverless car at Cruise unit

ft.com

58 pointscempaka6 comments

The headline is confusing. General Motors is “indefinitely suspending” work on Origin, but will focus their efforts on the next generation of the battery-run Chevrolet Bolt. Operations have resumed in Houston and Phoenix, and started in Dallas.

Ah, that makes a lot of sense, doesn't make much sense to go full custom for anything but PR until the unit economics are purely dominant over drivers, c.f. Waymo --- which is evaluating building their next platform on a custom car, but having been paying very close attention for years, you'll want to discount that, those sort of PR-focused partnerships tend to come and go. If you look back far enough you'll find a commitment to blitz-scaling via a prepurchase agreement for 100K+ vehicles, IIRC in 2017.

This is misleading. Here's the note from their official Q2 2024 shareholder letter:

"I also want to recognize the progress Cruise has made over the last several months. Our vision to transform mobility using autonomous technology is unchanged, and every mile traveled, and every simulation, brings us closer because Cruise is an AI-first company.

As you know, Cruise has returned to the road in Houston, Phoenix and Dallas and we recently made several significant leadership appointments, including hiring Marc Whitten as CEO. Marc has decades of experience on the frontlines of technology transformations.

The Cruise team will also simplify their path to scale by focusing their next autonomous vehicle on the next-generation Chevrolet Bolt, instead of the Origin. This addresses the regulatory uncertainty we faced with the Origin because of its unique design. In addition, per-unit costs will be much lower, which will help Cruise optimize its resources."

https://investor.gm.com/news-releases/news-release-details/q...

As a technologist who evaluates the evolution and viability of fundamental new tech on long time scales, the "reality gap" of most tech reporting on autonomous vehicles has been consistently terrible. Initially, expectations of near-term "full self-driving" were wildly over-optimistic. More recently, the tone has been more like "shit don't work." Yet neither extreme accurately conveys reality.

On one hand, the mythical "drive anything, anywhere, anytime" wonder-system a typical consumer would want, could use and could afford is likely still more than 10 years away. On the other hand, there are a ton of narrower use cases the tech is now finally good enough to solve at viable scale and cost - starting as early as next year. Unfortunately, we're now over-reacting into the typical 'trough of disillusionment' of the tech hype cycle just as things are finally ready to start getting really interesting.

https://archive.is/G8A40

At same time as Wayne gets additional investment? Quite a Divergence

Reply on news.ycombinator.com