[fastball]

> How are you proposing we assess the progress of the FSD team?

...any metric you want? Miles driven under FSD. Miles driven without intervention. Miles driven without accident. Anecdata from friends of yours who own a Tesla. Whether or not a partially supervised pilot program has been launched in some cities.

If Elon Musk said in 1999 "I think we will achieve self-driving next year", that also has no bearing on whether or not self-driving is achieved in 2025 (in either the positive or negative direction). It only means that Elon Musk's "predictions" can't be trusted as an accurate harbinger of success. Which is precisely why you look beyond his words and at the reality on the ground, which strongly indicates Tesla has made a huge amount of progress in the last 10 years, and could be very close to having unsupervised robotaxi service in various jurisdictions.

[cycomanic]

If we use kilometers driven with drive assist as a metric then nearly car manufacturers will have robotaxis by the end of the year.

If we talk about anecdotal evidence then I know people who are deeply familiar with the topic (working of self driving technology at other manufacturers) and they say fully self driving is still many years away for all manufacturers. Moreover the general industry sentiment is that Tesla is behind now and that more sensors then just cameras are needed.

But instead I should believe the Tesla fan boys who just like Musk have been raving about the amazing progress and telling me that FSD is just around the corner for years.

[fastball]

> If we use kilometers driven with drive assist as a metric then nearly car manufacturers will have robotaxis by the end of the year.

Sure, if you pretend that highway lane-keeping and universal A-to-B navigation are the same thing.

"What competitors say" is quite possibly the worst anecdata you could find as a broad rule, no? There is a wide gap between that and "Tesla fan boys".

[ImPostingOnHN]

Never heard of "universal A-to-B navigation", that sounds like google maps.

Is it fully self-driving, like Waymo? If not, then I'd lump it in with anything else that isn't fully self-driving. Either I can safely and legally nap while commuting or I cannot. Something that requires me to actively supervise the car and intervene as necessary is not self-driving, it is drive assist.

> "What competitors say" is quite possibly the worst anecdata you could find as a broad rule, no?

The post you're responding to is not simply repeating what competitors say, it is speaking of using data to avoid trusting what anybody says. Thus, this isn't a fair comparison. It should also be noted that you yourself suggested that the poster use anecdata.

That said, what tesla says about themselves is even worse than what tesla competitors say, if only because tesla is infamously untrustworthy, and their competitors are not.

But again: don't listen to what tesla et al say they will someday do, compare the data for tesla's drive assist vs tesla competitors' drive assist.

[fastball]

- I agree you should follow the data. That is the point I made from the beginning. Elon Musk / Tesla saying anything is not a sign of progress (or lack thereof).

- No, Teslas are not fully self-driving like Waymo. Alas, Waymo isn't the only competitor. The parent comment I replied to was lumping all driving assist together, as if Ford BlueCruise, which is highway-only driving, is comparable to current iterations of Tesla FSD, which has the capability to take you from point A to point B without a geofence ("universal A-to-B navigation") and with zero driver intervention required. That includes the ability to handle traffic lights, stop signs, roundabouts, pedestrians and cyclists, etc. Basically none of the other driver assists claim that capability (besides Waymo), and Waymo very notably has not allowed their cars to drive on highways in the majority of jurisdictions in which they operate (I believe LA is the only highway driving being done at the moment for Waymo). Tesla FSD however remains unreliable, which is why they haven't launched a full driverless service like Waymo, but Tesla FSD has more (unreliable) capabilities than any of their competitors, Waymo included. Reliability is super important though, which is why I'd say Waymo is clearly ahead.

- Not all anecdotal evidence is created equal. I suggested a specific type of anecdata – asking a friend (read: someone you trust who you think has broadly good judgement) who actually has a Tesla with the FSD package a question like "have you seen significant improvement in the self-driving capabilities of this car?" or "have you had any drives where your Tesla took your from home to where you needed to go, with zero interventions on your part?". Asking a GM employee "is your competitor doing a better job than you" is a very different type of anecdata and not something I personally would rely on. Mind you, I think asking a friend about their personal experience isn't great either, and that the other measures I suggested are much better, but it is still much better than "competitors say".

[ImPostingOnHN]

> The parent comment I replied to was lumping all driving assist together, as if Ford BlueCruise, which is highway-only driving, is comparable to current iterations of Tesla FSD...

You're saying "as if X is comparable" as if it is not comparable, but I and the other poster are saying it is comparable.

> ...which has the capability to take you from point A to point B without a geofence ("universal A-to-B navigation") and with zero driver intervention required.

Tesla "FSD" can absolutely, categorically, not do that. It requires the driver to constantly pay attention, to supervise, to mind the car, and importantly, to intervene when "FSD" fails. As I understand it, tesla will literally ban you from the "FSD" feature if you actually use it as FSD (hands off wheel, etc).

> Tesla FSD has more (unreliable) capabilities than any of their competitors

Unless it can actually fully self-drive (read: not require anything from the passenger), that just makes it a slightly more glorified lane assist, in my eyes.

> I suggested a specific type of anecdata

The request was over-specified: A tesla owner is far less likely to provide accurate, unbiased anecdata than an employee of a competitor, so I can see why the other poster volunteered the latter -- it is much better than "tesla owner says".

[fastball]

You can also compare apples and oranges – not sure what your point is.

Tesla FSD absolutely, categorically, can do that, assuming we agree that "intervention" means "take over for the car". Teslas are driving people from their starting point to their destination without any driver intervention, every single day. Are you being willfully obtuse when it comes to the distinction I am making between capability and consistency? Yes, you need to have both for a robotaxi service or to call the vehicle truly "driverless". If you don't have that reliability/consistency, it is indeed reckless to not have someone in the driver's seat.

If you really can't see the distinction, your eyes are blind. It's the difference between being driven by a toddler and by your alcoholic uncle. Your uncle has the capability, but he is frequently drunk so you wouldn't trust him to be your chauffeur. A toddler doesn't have the capability in the first place. Sober up your uncle though and the world is your oyster. But the toddler needs to do a lot of growing before it'll get you anywhere.

> A tesla owner is far less likely to provide accurate, unbiased anecdata than an employee of a competitor, so I can see why the other poster volunteered the latter

This is silly. I wouldn't apply that standard to anything else, and I doubt you would either. Who would you trust more to give you honest feedback about a Vitamix blender – your friend with good judgement who owns one, or an employee at KitchenAid?