I don't think your anecdote lends any evidence to the claim they're not close. Several reasons:

- You weren't using the actual FSD software, but the live production autopilot, so it's Apples to Oranges.

- It's not clear that the transition from low quality to high quality full self driving will be linear. Most claims around anecdotes come with an unspoken assumption we ought to see steady improvement over time vs a step-function like behavior. The latter seems somewhat possible given the nature of how improvements in AI systems seem to happen.

- Beyond all this, without understanding the underlying cause of the behavior you mention, it's hard to say from a risk perspective if the car made the right choice. Unless you have a clear understanding of the data and mechanisms that led to the speed reduction, it's hard to be sure it's a bug beyond not doing such transitions in a way that are comfortable for passengers. (My understanding is adding smoothing to the control algorithm is somewhat separate from overall decisionmaking, and hence poor control algorithms may mislead you into thinking it's more dangerous.)

Hullo, another Model 3 driver here.

I mentioned this in another comment, but it's relevant here; OP wasn't using Autosteer or Autopilot, they were using cruise control: a miles more simple system--and that system behaved in a way that seriously reduced confidence in it.

(For the record, I've experienced exactly the same behavior in my M3 as well.)

So, if we can't trust something as simple as cruise control, and we've seen similar confidence-eroding issues in the autopilot features Tesla has already deemed "good enough" for a live release, I think it is 100% reasonable to be skeptical that Tesla is anywhere close to "solving self-driving."

They have a history of overblown claims and faulty software in this realm--as much as I adore my car and the disruption Tesla has done to the industry, Tesla does not deserve the benefit of the doubt in this realm, especially not with safety-critical systems.

Cruise control is exactly that, simple. It errs on the side of safety and makes fail-safe mistakes.

FSD is a completely separate system with different tolerances. One does not have any relation to the other

The transition from low quality to high quality driving will be sublinear, not linear or greater than linear. This has been the case at every other self driving company. You get big gains from working on low hanging fruit early on in the process, then there's a long, arduous march to make improvements once those fruit are gone and avoid regressions. That's why Waymo, Cruise, Zoox, etc. Have been able to do impressive demos for the last decade, but it took Waymo a full decade to actually launch a public ride-sharing service.

Before you bring up Tesla's "Fleet Learning", it's not used at all outside of gathering perception training data for labelers. greentheonly got access to a dev build of the FSD beta - ML is just used to generate a representation of the current scene - all of the driving and planning logic is rules-based and hardcoded by engineers.

Seems unknowable still since Waymo and Cruise are now both deploying fully autonomous vehicles ahead of most peoples' expectations. If graphed, I'd imagine this implies a non-linear curve given these expectations were grounded in linearity imo. So it's under-determined basically.

They're deploying AVs way (1-2 years, at least) behind their own schedules, by their own admission. Ask anyone inside of the industry (including Musk) and they'll tell you that the problem is the long march of nines I mentioned earlier.

This is also why you saw a bunch of AV startups with < 50 employees that got big initial funding rounds but folded after a couple of years - it's pretty easy to grab a few dozen grad students, slap some LIDAR on a car, encode basic road rules, and get a demo working that's compelling to investors. Once you get past that initial pitch though, you get into the weeds of trying to progress without causing regressions elsewhere.

Also, what makes you think those expectations were grounded in linearity? Waymo went from an initial project at Google in early 2009 to driving 10 complex 100 mile routes without intervention in mid 2010 - on the surface, it would have seemed they were 95% of the way there. Improvement has been sublinear for a very long time and everyone in the AV industry is aware of it.

The linear assumption I'm referring to is forwards-looking not backwards-looking. Up until say last year I think the consensus was that a production-deployed FSD vehicle was ~5 years away. It shipped this year for both Waymo and Cruise. Unless you're arguing that the pessimists and optimists have just been arguing over the slope of the line, not the curvature (which may be fair.) If that surprise is any indication it leads me to think a similar consensus-breaking event may happen with Tesla, but time will tell when that is, since as you state they are behind in several ways.

Cruise and Waymo Gamble Lives with Unsupervised Self-Driving Cars

https://jperla.medium.com/tesla-saves-lives-waymo-gambles-pa...

Also a Tesla owner, on latest AP hardware: I would be amazed if the FSD update manages to solve all the problems, I have been through a few updates where things noticeably improved but we now have to pay Tesla money because while using autopark it just completely reversed itself (at a pretty high speed) into an object. It was insane. We’ve had weird AP issues too and now we don’t feel confident using any of those features.

I agree Tesla's shipped software that doesn't inspire confidence in the current version. I only use it on highways and remain fully engaged. (If I had to guess, in that scenario it does provide a marginal improvement in safety over fully manual driving.)

However I don't think lack of confidence in the current software, particularly the production autopilot software, is really giving me much signal in helping predict if Tesla will have FSD anytime soon or not. Now that FSD (at least in Phoenix conditions) is known to be possible the prior that it will never be achieved by Tesla to me is near zero, beyond the chance that LIDAR is strictly necessary and they have a doomed hardware stack. I think there is a pretty big range of possible timelines though - I could see it happening in as soon as 18 months or as late as 10 years depending on how poorly things sift out with their hardware stack.

Can you fill me in on where the levels of Tesla software are? The version of the OP who you are calling out as "apples" (the old, legacy one), what SAE autonomous level is that one? And the "oranges" one that is new and beta, is that a different level? If so, which level?

* Level 0 (No Driving Automation)

* Level 1 (Driver Assistance)

* Level 2 (Partial Driving Automation)

* Level 3 (Conditional Driving Automation)

* Level 4 (High Driving Automation)

* Level 5 (Full Driving Automation)

Different features of Tesla's software have different levels. As a whole, it's level 2.

If you restrict it to certain parts of driving, eg TACC, that might be level 4 (doesn't conform to the chart though). It can handle everything in it's domain, albeit in a rudimentary way, so a human doesn't always have to shadow it, but they do need to be ready to take over in some cases.

The apples to oranges bit is using the cruise control software to make guesses about FSD. I'm not certain but it would be very surprising if the software that provides basic cruise control shares much of the underlying algorithms and models as FSD, and if it does I'd be surprised if it's using models that aren't quite a bit older.

> It's not clear that the transition from low quality to high quality full self driving will be linear. Most claims around anecdotes come with an unspoken assumption we ought to see steady improvement over time vs a step-function like behavior. The latter seems somewhat possible given the nature of how improvements in AI systems seem to happen.

That is true, but it's actually an argument _against_ the assumption of full self-driving capability being just around the corner. Because that assumption itself is based solely on observed recent progress, which - as you correcly state - can be very misleading when used as a means to extrapolate the future due to the tendency of the AI field to have long stretches of very little to nearly no progress at all, divided by huge leaps followed by stretches of very fast progress in specific sub-areas of the field. We clearly have seen such a stretch recently, triggered by the deep learning idea and enabled by an explosion in parallel computing capability over the course of the last decade.

I'm not claiming it will be linear or will not be linear. I'm saying it's a bad idea to assume either when trying to predict the future.

The problem is: you have to assume "something", because otherwise you simply can't make any predictions. Either assume progress is linear (which seems to be the instinctive default most people use if they don't know much about a particular field) or assume it's non-linear. In case of the latter, you can choose between assuming exponential progress to apply or a step-wise, non-continuous kind of progress. It's a bad idea to not assume one of these to be the case if you want to make a prediction, because it's not possible to predict anything without that assumption.

I mean, you can of course refrain from making any predictions. But I have never met a single person who could resist the instinctive urge to extrapolate current developments into the future. It is as much a core part of being human to do this as it is to breathe air.

It's the assumption of linearity in combination with anecdotes that make the prediction weak, not the assumption itself. If you hold that assumption and use a variety of data to justify it and extrapolate from that's different than what I was criticizing.

"You weren't using the actual FSD software, but the live production autopilot, so it's Apples to Oranges."

Do you think Tesla has completely separate groups of engineers working on this? That doesn't seem like Tesla's style at all.

It's the same people but FSD was a rewrite of the system.

The rewrite was to do image recognition and inference not from individual still images (pre-FSD) but from video feed from all cameras. This was enabled by the FSD chip Tesla started putting in cars about a year ago.

Plus the improvements that come simply from gathering more data from cars and using it to improve the training of neural network.

It's documented in Musk's tweets and Karpathy's talks.

> Beyond all this, without understanding the underlying cause of the behavior you mention, it's hard to say from a risk perspective if the car made the right choice.

What argument could there be that suddenly slowing down on the freeway with no visible obstruction is the right choice?

It does not avoid any danger, but it creates a danger of getting rear-ended by cars behind you.

The argument would be: the car saw something you didn't.

I'm not saying that's what happened, just that it could have happened and therefore would be an explanation.

Your position here is, essentially, the same as a Q believer. Just trust in Elon, wait for the big reveal. Despite a complete lack of evidence in favor of the idea, and also despite an abundance of evidence to the contrary, FSD will emerge fully formed in the near future, without any intermediate signs of progress.

You have a reading comprehension problem. Saying "what you just said does not confer evidence to your claim" (which is what I said) is not the same thing as saying "I am 100% certain of the counterclaim." Responses like this one are why discourse on the Internet has fallen so far. The analogy to Q lands you squarely in the "lol fag" tier of responses.

I think phantom braking with TACC is a function of it being alpha software. With the beta stuff (everything else), the driver is expected to always be alert and intervene where appropriate, so Tesla doesn't have to be as aggressive on the safety side.

As it relates to the features, I agree that they aren't ready for what most drivers expect, either because they aren't reliable enough (beta) it because they aren't confident enough (TACC). With that said, FSD is a feature set, not a level of reliability. As a whole it eventually needs to reach a level of reliability, but there's no specific time horizon.

For Tesla, a feature moving from beta to alpha is what indicates reliability. In addition, that reliability may not come with the degree of confidence most people expect from a driver. Safe FSD could just end up as the AI equivalent to a student driver. It'll get you someplace safely, but there may be phantom braking, excessive waiting, and so on. If that's the case it would be nice if we could enable a beta version with more "normal" driving that we would have to keep an eye on.

The FSD update which merges the cameras together inside the NN is so drastically different than the software you're describing your experience.

You might as well be comparing your experience driving a BMW or Cadillac with SuperCruise. Your experience on the old tech stack is irrelevant.

Well, apparently it's the same NNs. The FSD beta just has more stuff turned on:

https://twitter.com/greentheonly/status/1343670544529575936?...

https://twitter.com/greentheonly/status/1336467014727110656?...

The guy you're quoting gets his information from spelunking in Tesla firmware. He's not a dev on AutoPilot team.

It's well documented that FSD is a architectural rewrite of the code.

Pre-FSD the input to NN were individual still pictures taken by the cameras.

New architecture feeds NN with video feed from all cameras.

New architecture is possible thanks to new FSD chip that is 20x better that was Tesla used before.

I don't see what definition of NN would consider them the same system.

On YouTube you can find talks by Karpathy where he goes into detail about this stuff.

> Your experience on the old tech stack is irrelevant.

It's cruise control. If you couldn't fry an egg without burning it, forgive me for being a little skeptical when you claim that you can make award-winning souffles after having attended some cooking classes.

That's quite reassuring to hear. It's surprising to me that the software wouldn't have been stereoscopically merging cameras already, but adding that capability does sound like the kind of thing that could actually lead to an entirely different level of performance.

There is a history of overpromising that will be still leave me somewhat skeptical until I see it. But it does sound like a move in the right direction.

Perhaps, but if they can't get an orders of magnitude simpler adaptive cruise control right (not just right, but perfect), it doesn't lend much confidence that they'll get FSD right any time soon.

The protestations to the contrary here reek of defensive fanboy/apologist behavior.

You can watch videos on YouTube of FSD beta doing very complex things.

It's clearly order of magnitude more advanced and capable than pre-FSD AutoPilot.

Calling anyone who disagrees with you a "fanboy/apologist" does the opposite of strengthening your argument.

Fellow Model 3 driver here.

> You might as well be comparing your experience driving a BMW or Cadillac ....

But OP isn't--he's comparing the current live Tesla features with paywall'd / betawall'd features on the same car, from the same manufacturer. I cannot think of a more relevant comparison! I think it's completely reasonable to doubt Tesla's (and Elon's) claims in this realm, when the existing, public examples of this technology fall far, far short of the kind of confidence we'd want to have in these systems.

> Your experience on the old tech stack is irrelevant.

I 100% disagree. If Tesla can't even get a basic function like cruise control to work well enough that drivers are confident in it, why would I trust that they can get self drving to work well? Especially following things like the Summon beta, or similar fiascos?

I still get twitchy when my M3 comes up on an off ramp it's not supposed to take on autosteer--I've had the same behavior as OP (sudden unexpected speed decrease) happen far too many times for me to have confidence in that system. In turn, I have even less faith in the more complicated systems, like autosteer or FSD.

“I love my Tesla but” has become so overused it is on the verge of becoming a meme.

So only unconditional love for something or someone is not a meme?

no, there’s just a fascinating phenomenon where Tesla owners appear to shrug off bad quality control, bad service, buggy software and controls, safety issues, etc en masse because they love their car.

I guess I'll take the bait here.

I have owned 2 cars that I bought the first year the model came out.

The Model 3 I have had a few dumb issues like, if you run the paint app sometimes the gps location on the map doesn't update until you close the paint app and your displayed location lerps from the old point to the new one over a minute or so.

The Mini had an eco mode that would occasionally shut off the engine entirely at highway speed and not allow restart until you came to a full stop and put it in park. Eventually, they just disabled eco mode in software Which took a few weeks of me working from the mini dealer. There was display of how many miles you had gained that reset when you filled the gas tank. My record is 42 Miles gained in eco. After the final update to the software, which was a full day in the garage, My best was 8.

One of those was worth running stories in the press about.

I guess what I'm saying here is that there is a deep vein of shadenfreude in the news about Tesla. A single Tesla on fire is a big national and some times international news story. I have stopped even reading the damn things. It seems like a weird combination of them being seen as rich people cars, early adopter cars, and liberal hippy cars. Everyone can hate someone on the spectrum. Typically, the issue in articles about Tesla flaws is so isolated and dumb that it exists solely to generate clicks. The last article about Tesla I clicked through stated that consumer reports had downgraded their durability rating due to the discovery of single human hair in the paint of a model Y. I get that it sucks. But, I don't get why I should care about it.

I own a Model 3 because I needed a four door car because I had kids, electric seemed like a net win, the local Bolt dealer told me not to buy their car, and the range let me drive to the grandparents houses without a charge. I recognize that there is stuff about the model 3 and Tesla that sucks but overall it has been a pretty solid car for the last 2 years for me.

Would I recommend one to you? Yeah. I don't think there is a company that offers the combination of electric car, mature tech, range, and charging network at a similar price tag.

I think it happens in just about every realm when you get close to the innovative edge, especially in tech: you can be really excited about some aspects of $SHINY_NEW_THING while recognizing that it has serious drawbacks in other realms--possibly realms that people expect $SHINY_NEW_THING to excel at as a contender in the space.

(inb4 "Sure it goes fast, but the panel gaps!")

Personally I love Tesla-the-company and my Tesla vehicle because of the innovation it's driving in the space--and I'm willing to tolerate the ways it falls short due to how incredible it is in other ways.... but I do try to be up front about those shortcomings when I discuss it, at the risk of being one of those "I love my Tesla, but..." people.

The car is comfortable, responsive, stylish, fun, and feels great to drive. The cruise control is fantastic in stop-and-go traffic. I always look forward to driving it. And then every once in a while does something that drives me absolutely crazy, for a minute, and then goes back to being fun again.

the biggest issue is the service model not anything related to the car unless you got a lemon.

they aren’t scaling service centers fast enough to keep up with an extra 200k cars added on the road every quarter, so service meaningfully deteriorates every week as more cars are sold and existing cars age. it simply doesn’t work unless they were committed to building out 2x-3x the SCs they have now but nothing indicates that’s on the roadmap at all.

Here's a Tesla 2020 Q3 investor update https://tesla-cdn.thron.com/static/4E7BR9_TSLA_Q3_2020_Updat...

The numbers:

- increased service locations 12% YoY - increased mobile service fleet 8% YoY - increased supercharging stations and connectors 32% YoY

They are clearly building more service.

During quarterly conference call Musk stated that he believes service is the biggest driver of sales and that Tesla will invest in that. So contrary to what you claim it's on the roadmap.

Many people are over the moon with Tesla service given that mobile service comes to you and they can do fixes without you even there. That's something no other car maker does at scale, as far as I know.

Did you have a personal experience with Tesla's bad service or is it just amplifying FUD from other people?

growing service centers 12% YoY when they added 50% of cars they’ve ever sold this year alone and plan to sell close to 1M next year (100% YoY) is the definition of inadequate.

Every car is a game of tradeoffs. Any other EV you can buy today (including the Mach-E) will have you trading off build quality and access to good service (probably; some dealerships suck at servicing EVs) for extremely slow software, low range, and lack of supercharging access. MKBHD has a good video on this that came out yesterday - https://youtu.be/pJlA2J92Fw4

Check judge2020's reply to bmcahren - you have the same software running, the FSD beta just uses more features from it.