Hacker News Reader

Comment by Animats | Hacker News Reader

The comments from Apple indicate that what they're really concerned about is disclosure to the media, not someone copying their self-driving technology, such as it is. Apple put 5000 people on this project and failed. What they don't need is an insider writing an "Apple's Biggest Failure" book. I wonder if we'll ever see a technical post-mortem on this.

View on HN

>Apple put 5000 people on this project and failed.

not sure. These are still driving around their Sunnyvale campus:

https://www.macrumors.com/2017/08/25/apple-new-autonomous-dr...

Btw, one of the best sensor suites around (except for Google who seems to have reached kind of optimization stage where they started to remove the "extra" sensors). And Apple has highest number of self-driving permits in CA if i remember correctly. Though i don't see them in "disengagement reports" (https://www.dmv.ca.gov/portal/dmv/detail/vr/autonomous/disen... - interesting reading, Google's average human driver reaction time was 0.91s)

I would be really interested in an Apple designed car, not exactly an Apple created self-driving tech suite. I trust Apple's design far more than I trust Apple's hard tech chops, especially when it comes to processing massive amounts of data

I wonder about the assumption in that MacRumors article that it's a self-driving car, given this article about Apple building their own Maps data set: https://techcrunch.com/2018/06/29/apple-is-rebuilding-maps-f...

Given that the MacRumors article says that the Lexus was parked, it might be a human-driven car for gathering maps data?

the quick way to tell between self-driving and mapping car sensor suites - density of the point cloud it collects, ie. type/number of sensors with the self-driving is unnecessary dense for mapping and their directionality - the self-driving is more forward looking as well as immediate vicinity of the car while the mapping is sideways and doesn't care about the immediate vicinity. I'd speculate that Apple, historically being weaker when it comes to [higher level] software, seems to have issues with building of dynamic scene perception from the big can Velodyne stream and are trying to solve it "hardware way" by structuring the perception data in more time/space consistent way inside limited sectors/planes.

The head of Google's self driving program has publicly stated that fully autonomous cars that can handle challenging conditions are decades away.

https://web.archive.org/web/20160422140357/https://spectrum....

Does that make Google's program a failure?

Apple decided it didn't make sense to work on building a car when the software to control such a car isn't close to being there, and moved their focus to software.

In the last keynote they mentioned that they have solved some snow problems: https://youtu.be/ogfYd705cRs?t=6177 :)

That's Chris Urmson's talk at SXSW in 2016 [1], which has been widely viewed. It's one of the best talks on the subject. He said "How quickly can we get this into people's hands? If you read the papers, you see maybe it's three years, maybe it's thirty years. And I am here to tell you that honestly, it's a bit of both." Then he showed video of Google cars dealing with various challenging conditions. This included the famous scene where a Google self driving car encountered a woman in a powered wheelchair using a broom to chase a turkey. The software recognized this as something to avoid, even though it didn't classify it further.

[1] https://www.youtube.com/watch?v=Uj-rK8V-rik

The challenges Google thinks will take decades to overcome are those posed by streets that aren't new, extra wide, well marked, and with low traffic in an area that almost always has excellent weather conditions.

Basically, they have a solution that they are willing to put into limited testing in certain parts of Arizona, just as Apple has decided to do limited testing with autonomous employee shuttle vans between their campuses.

So Google has moved from "this is our moonshot program that we will solve in a couple of years" to "this will take decades of incremental progress".

Given that Google and Apple have both come to the concision that this is a long range incremental software problem, and given that you say that this represents an enormous failure on the part of Apple, how is this not also an enormous failure on the part of Google?

Failed in what sense. The project is on, they reduced headcount. Maybe they don't need the same amount of people they initially they needed or they changed the goal of the project. I would hardly call it failed. Other companies have been working longer with a bigger team without a self-driving product to show for it.

Only Apple could have a round of layoffs and nothing to show for a project and not only will people not call it a failure they'll even act surprised at the suggestion of failure.

I am not an apple fanboy, you can read my comment history. People can have different opinions without being a fanboy.

I...don't follow. Are you claiming to be an Apple insider with knowledge of said project and can prove it actually did fail? Without that I'm not sure how you can say it was a failure or not from the outside. You literally know almost nothing about the project.

I could think of dozens of reasons one might do this. Sure they'd likely be wrong but I don't understand the view point of "it's obviously a failure but everyone is giving Apple a pass".

Apple fandom is amusing. In the early iPhone days, when they had screen scratch problems, I said that they should use a sapphire coating, like the checkout scanners at Home Depot. Those survive steel tools being dragged across them all day for years. Fanboys insisted that sapphire was unnecessary. "You don't need that much scratch resistance". "Costs too much". Etc.

Then Apple announced they were going to put a sapphire coating on their screen glass. Fanboys then insisted that sapphire was the Next Big Thing, and phones without it were inferior.[1]

Then Apple's sapphire scheme fell through, after they pressured their sapphire supplier into contract terms that put the company out of business. Apple went back to Corning's hardened glass. Fanboys followed the company line and insisted sapphire was unnecessary and would crack.[2]

Kyocera and HTC make phones with sapphire screens, but Apple fans don't talk about that.

[1] https://www.cultofmac.com/267068/everything-wanted-know-sapp... [2] https://www.recode.net/2016/7/28/12305062/apple-iphone-goril...

Apple animosity also amusing, just because you have an idea, it might not be implementable for various reasons. Expressing those opinions are fine however.

Calling people fanboys only discredits you. We don't have a single opinion, you've just heard different groups of people. Sapphire has advantages and disadvantages, it's more scratch resistant but also more brittle, for example. Would it be nice?... sure. Would it be a deal breaker? Definitely not.

I think the issue is on who decides it’s a failure.

If it’s a consumer goods we choose to buy or not, we have a yardstick to decide (i.e. does its utility match the price paid?).

For an internal project not even acknowledged in the broad, what’s the yardstick? The expectations we have based on our imaginations and dreams about the potential product and its timeline ?

Perhaps Apple sees it as failing the original goal, but really who knows ? (and to be honest why do we care what Apple thinks about it)

I'm far from an Apple fanboy, but Apple, Google & co have so much spare cash that they can go for 5-10 large moonshot projects and still not care about these "failures". Nokia betting on Symbian, Microsoft on Windows Phone, IBM on OS/2, those are failures, things which put the company in jeopardy, in my opinion. Everything else is just "research" :)

Reply on news.ycombinator.com