12 Comments
тна Return to thread
Comment removed
Aug 16, 2023Edited
Comment removed
Expand full comment

Considering the smart safety features (lane assist, adaptive cruise control and brake) on my Subaru Ascent turn off during a heavy rain and can act squirrely in dry conditions, I'm certainly not going to trust a fully-autonomous vehicle to handle anything.

Expand full comment

I think this bar is too high. If you have a car that can handle 100% of driving conditions in Phoenix then for that purpose it's a successful self-driving car. Not in MN.

Seems obvious that self-driving will improve and expand out from the very limited places where it works today.

Expand full comment

How do you disable it when its owner tries to drive it to Brooklyn?

Expand full comment

That seems like the easiest problem in the world given that self-driving cars use their own MAPS to decide what route to take. The car knows where/when it is currently able/allowed to self-drive; that gradually expands over time as capabilities improve. You buy/lease/borrow the car if its current/likely capabilities meet your needs; if not, not.

If some vendorтАЩs technology canтАЩt reliably handle Boston or Brooklyn winter driving yet, the car can warn you to that effect when you try to go there and automatic routing will be limited when you arrive.

Expand full comment

I was thinking more of the fact that autonomous vehicles can't handle merging on the BQE. How does it figure out when the drivers are too aggressive and the roads are too congested?

Expand full comment

What makes you think they "can't handle" merging specifically there? You seem to be treating this as an established fact but I'm not sure where that fact comes from. Merging is hard for people too - both autonomous and human-driven cars will sometimes be awkward or slow to enter congested areas.

Expand full comment

Google it.

Expand full comment

I did google it. I found an article or two in places like _Wired_ saying merging was a challenge - duh - but I found nothing to suggest they literally *can't* handle it and it's not improving nor anything to suggest the BQE in particular was a problem. That's why I asked YOU. Since you're the one making this bold flat assertion I figured you might have some SPECIFIC source for the information.

(Saying "Google it" is rarely helpful anyway because Google tailors its results to the person searching so what YOU see when you google something is different from what I see when I make the same query. Not to mention that we're likely to be making different queries since you know what result you're EXPECTING to find and I don't.)

Expand full comment

That article doesn't say self-driving cars can't handle merging on the BQE. Heck, it doesn't MENTION merging nor the BQE. It's just a generally skeptical take from last year largely on the BUSINESS end of self-driving.

Takes of that sort should always be expected as it's *hard* to launch a new industry. Early in the development of any new technology, many companies will think they MIGHT have a competitive edge and be motivated to invest in doing it themselves but over time most will discover that they didn't have the edge they thought or will hit snags in development and drop out, expecting that they can ultimately BUY or license the tech from the eventual winners. Meanwhile those who stay in the fight will almost always find themselves pivoting and trying various strategies to get off the ground because nobody yet knows exactly where the most or easiest profit is to be found.

(The silliest quote in the article is (as an example of "Industry leaders backing off on big claims"): "Waymo has said it will be decades before the technology is everywhere in the world." That is not "backing off"! Did anyone actually claim it would be EVERYWHERE IN THE WORLD in less than "decades"? If so, that was a silly person who we shouldn't be letting set the bar for what constitutes success. Lots of people think it will be popular - maybe even common/ubiquitous in *some areas* - but it's a long way from that to EVERYWHERE IN THE WORLD.)

This month, Cruise greatly expanded their service volume in SF, increased by 20x the size of the region they serve in Phoenix,AZ, and started test service in Atlanta and Nashville. Expect to see more articles like this:

https://sfstandard.com/2023/08/11/san-francisco-robotaxi-cruise-debauchery/

Expand full comment

If you think that article is just talking about the business end rather than the technical aspects then you need to go back and reread that article.

Expand full comment