Search results

I seriously doubt emailing adds any value to the feedback loop. Explaining what didn't happen, should have happened, etc. is not something a console operator can really influence, esp. if we are talking about collecting mass video training sets. Asking a human to sit in the middle to moderate...

At some point you'll either have to update or have your car neutered. https://electrek.co/2020/02/14/tesla-force-software-updates-owners-resisting-them/ FSD beta is a different beast than the main line builds. There is nothing but upside for 2021.40.x on the non-beta side.

With training sets, as long as they can be classified, more is always better (at this stage). We are extremely early days here... regular AP is based on billions of miles driven. I doubt they have more than 100MM miles of FSD on the books. For instance, they definitely don't have enough capture...

A big piece of driving is about showing intent through action. We drive in a way that says: "I intend to make this turn (and you should yield to me)", or "I am not moving (so I am yielding to you)". Accidents happen when that predictability goes away. We get frustrated at drivers that are doing...

They need a lot more drivers and data to be successful. I can't see a scenario where the beta would be too big at this point ... only too big, too soon.

65-30 would be unusual. I've seen it scrub 10-15mph at highway speeds (e.g. 60+ mph), more akin to you suddenly lifting off the accelerator. It feels more severe because you're not anticipating it (same reason passengers tend to be more susceptible to getting car sick while you're fine as a...

I believe the "request FSD" button came with 2021.32.22. So until you're at a current, non-factory production build (likely 40.6), you won't see it. Once you do, then you can opt-in and play the Safety Score game. You cannot request an update to a newer version - you have to wait until Tesla...

I'm certainly one of the ones who got upsold too hard on the good. My first 1000 ft on day 1 was the worst, but it definitely set appropriate expectations and spooked me good. I think had it been smooth and incident-free, I might have become complacent too soon, because sometimes it's actually...

That's what I would expect... if 2 nodes cannot agree, the tie-breaker should be to abort. It's definitely the safest outcome.

I agree with this long term, but it assumes near and far depth are able to be computed well by vision. Karparthy demonstrates some very nice depth mapping from a car's immediate surroundings, but I remain unconvinced for long range, where there is only 1 camera, which previously could lean in on...

Thanks... this is interesting to me. Typically you see redundant architectures work in parallel and sometimes by consensus. You see that esp. at the chip level as EPEs become more relevant with chip densities increase. Also seems odd that the B-side is asleep until the A side fails, esp. since A...

Lidar and radar is superior to vision-only for obstacle detection in 2021. Prove me wrong.

Not comparing who's ahead or not. Just pointing out that it clearly works. Only the Tesla religion says radar and lidar are wrong. The reality is radar and lidar are both perfectly serviceable for certain purposes. For stuff like phantom braking, lidar > radar > vision all the time. Using 2 or...

Unlikely unless they can trust the interior camera. 10.5 was the first pass at this, so probably a few iterations to go.

I did Waymo in AZ a few months back. It uses radar and lidar. It works. It doesn't solve Tesla's problem, but it does solve a real-world autonomy problem very well, without a driver. You can drive a Tesla everywhere in the US with FSD beta, but always with a driver.

I never said infallible. We're still beholden to what was coded (or not coded). As it stands today, Vision is behind the curve. I suspect we're going to get there in another year. The transition away from MobileEye was fairly quick as well. For now, Tesla isn't ready to do a follow distance of...

There is no singular correct approach. But we do know that what they're doing works quite well. I can hail a Waymo today with no driver and enjoy a true driverless ride, albeit with huge geographic caveats. Tesla is trying to solve a much broader problem, so yes... vision is needed to solve for...

The more points of truth that exist for the vehicle to make a decision, the more confident it will be. Radar is not inherently more useful. However it provides a completely different perspective on the world that is a completely separate data source, which is extremely useful. This is also why...

Oddly convenient that the removal of radar and the push for pure vision came at a time when there was a radar shortage. Meanwhile at that time S/X cars continued to get radar installed. Tesla is choosing a novel approach, but the benefits of radar are well known, and in service today in the...

Even if s/he's wrong. 😂 I don't disagree with some of @EVNow's points... but I hold firm that an experienced driver in San Jose, CA that has never driven in Boston would question their driving abilities in Boston. It's very different, and that also carries through to autonomy.

Radar adds noise, but is very useful as another data point. It's not an accident that radar + vision has faced less overall phantom braking. All I'm saying is at a distance, with Tesla Vision, there is 1 source of truth with a single eye which cannot judge depth "well" at a distance, but can...

There is also a world where FSD beta is opened up but with more confirmations. Like confirm to change lane, hold turn signal to complete turn, confirm to proceed on green, etc. Not saying that's what they'll do, but a lot of the risks could be mitigated allowing it to become more available...

It can see objects further at high speeds, but it cannot easily determine depth. I'd guess this is where radar really helped... there were 2 sources of truth, and the car used these two sources to decide whether to brake or not. Stuck with long range monovision, you don't have that advantage.

If cities weren't materially different beasts, a new driver would navigate them all with ease. It's the combination of road dynamics, driving habits of others, casually marked signage, local driving norms (which do not necessarily follow any legal norms), and a certain level of driving acumen...

My theory would be more costly (processing power + time), and not useful for the near range depth mapping. Probably also harder to create binocular vision using the narrow since it looks so far out. As humans, we don't really perceive much depth at a distance.

Same but different. We need localization or we will continue to have to disengage. It's not done until FSD takes a masters class! I'm closely watching the Waymo testing here in NYC. Was very impressed with my experience with them in AZ, esp. the parking lot drop-off. Want to see if it feels...

This is the curse of autonomy - it's following the rules correctly, but as a human we recognize that double rapid lane change requires you to make a judgment call on whether there's enough leeway at that moment to do it. And even from the left, you could probably make it, but FSD is not...

I guess this is the nuance... there's enough little things that don't happen frequently elsewhere and are not known problems that can be solved for with the software today. NYC has their own, just as Boston, SF, DC etc. do as well. To me, I see these as city specific models. Cues it does not...

Biggest ones that come to mind that I'm dealing with is what I can best describe as a funnel merge. The merges entering the Holland are a really good local example entering NYC... 4 and 5 lanes into 1 lane, which are not marked. FSD cannot handle this. It's not a standard zipper merge...

I am seen. 😂 It is different, even if the traffic controls are largely the same. We have some very complex scenarios that are unique to us, just as other cities have their own quirks. Certainly above 14th St, it's all grid so easy to navigate right and left, but obstructed view lefts and rights...

Oh yes, this also happened to me as well. I think it can be generalized that on some level, once you make a decision, you need to be committed to it. More acceleration is your safest move once you're "in the box". Lot more work to be done here.

That someone would be very wrong. Let's not downplay the reality FSD still chooses the wrong lane periodically. It can misinterpret traffic signals when there's a dedicated turning light. In addition, 10.5 seems to fail to signal in some cases. I've seen all 3 of these in the course of a day...

Creating new options to regress behavior is problematic. I would not support this. The car has always supported smart cruise control. Now we want to add another tier of behavior where for some people the car no longer slows down and goes into FCW mode if it charges into a traffic jam, but for...

Hardly. Pioneer maybe, but this was a proof-of-concept. Audi put it in production cars with working with supporting cities with working infrastructure. That’s a whole different game.

My theory: because people are resetting and doing whatever they can to get a 99 or 100, 98s will only have a slim chance of getting in. I would assume they have more available testers than they need at each new release, and so batches are getting inducted with newly minted 100s and 99s first...

You just need to create an account and then you can access the map. https://www.tomtom.com/mapshare/tools/new/mapshare/

I don't know whatever happened to GLOSA but Audi was ahead of the curve on this... https://www.theverge.com/2019/2/19/18229947/audi-traffic-light-sensor-green-wave-v2i

Wow, that's pretty epic.

Yeah the thing is both sides would have their phone records subpoenaed long before the cabin cameras is subpoenaed. It would really have to go to court before someone goes through a lengthy discovery.

My perspective is skewed since I bought EAP in 2018 and then got FSD when it was $5k. I *think* I'd buy FSD again, but I'd have to think hard about $10k vs monthly subscription.

Take a closer look at the front lens of the cabin camera. Might just have a little haze on it. I've had an obstructed side pillar warning in the sun when the car was dirty.

What you're suggesting here is "good thing nobody saw what I was actually doing when I killed the pedestrian". Again, I ask what will that camera capture that isn't going to be something that we should truly be liable for? There are a million reasons you'll get away with something without it...

Of course this can be argued for self-driving all around. It's not hard to drive, follow traffic, keep in a lane and pass as needed. But the reality is nobody drives at 100% attention, 100% of the time. The value in these systems isn't here to support laziness... it's to ensure that even when...

I would not be surprised to see it be required with standard AP sooner than later. What use case are you afraid of where "if both driver's did something stupid there will only be evidence against the Tesla driver"? What possible piece of incriminating evidence is this camera going to produce...

Interesting that you feel this way. From my experience, NoA and auto lane change are ridiculously well polished, and things I use all the time. Tapping the turn signal and knowing the car is going to nail that lane change is great - also highlights the blindspot car on the screen. NoA has been...

I understand what you're getting at. It's still an open door to people to abuse the system. Lot of people want Tesla to fail. Just a sample of what the "good guys" are doing trying to identify security issues: Hackers show how Tesla Autopilot can be tricked

Part of the reason this doesn't exist is because it invites all sorts of bad actors to do things with these stickers to limit the utility of it at best, and to create dangerous situations at worst. Having you back into a car because someone put a sticker on a random bumper would be just one of...

lol, this is the question I've even asked NYC officials many times. I've lived here for 20 yrs, and every few years, I try and find out because someone will do something crazy like 4 across and it'll bother me. Usually I see 2, and I assume one is for turning as you suggest. However, nobody will...

In my case, it just sat behind this line of cars on the right. The car actually started on the left, ping ponged left and right, but then saw cars lined up and moved over to the right to get in line. This street is definitely an edge case - you can see how wide it is. FSD is clearly looking for...

To CA's credit stuff generally seems to be well marked. And of course Tesla engineers have driven those roads far more as well, so they may as well be full high res mapped. FSD cannot handle my NYC street whatsoever, but I get it - it's a wide one-way street that people treat as 1, 2 or 3 lanes...