OFFICIAL BUTTON WATCH

[azred]

No, they must want to ensure it is as safe as possible; thus limited numbers of BETA testers initially.

Yes, this makes sense. Nothing says safe driving like having some YouTube entrepreneurs narrating videos while driving. BTW Elon claimed he revoked Beta status from those not safely driving. But I guess free PR on YouTube trumps safety.

 

[gearchruncher]

You can't fix stupid. People use beta versions without heeding all the cautionary statements.

I'm sure as a pharmaceutical law professor, you have examples where a cautionary statement was deemed insufficient to absolve the manufacturer of all liability. That, in fact, calling your users "stupid" is not a get out of liability free card.

Human Factors Engineering is basically the idea that you can in fact fix stupid with good design, because what we often call "stupid" is just human nature, and the products we design are meant to be used by real humans, not theoretical perfect humans that do not exist. A disclaimer is basically the worst, laziest, and last resort way to do HFE.

 

[Daniel in SD]

You can't fix stupid. People use beta versions without heeding all the cautionary statements.

Let's see what beta 9 looks like.

I assume they're not waiting for people to get less stupid.
What might we see from beta 9 that will make release more likely?

 

[qdeathstar]

View attachment 682967

Seriously? A pharmaceutical lawyer/professor upvoting the idea that if no injury has yet occurred with a product, it must be safe? Pharmaceuticals are the poster child for unexpected consequences, hard trade offs, ethical missteps, and pulled products.

Pseudoaddiction


Tesla and space x are not safety first companies.

 

[Cheburashka]

You can't fix stupid. People use beta versions without heeding all the cautionary statements.

Let's see what beta 9 looks like.

I don't think beta means what Tesla thinks it means. I'm pretty sure AP1 cars are still technically "beta"? I'm pretty sure that AP in my "FSD" car still has the beta label too...

 

[dgatwood]

However, if you think you can estimate your level of exposure from 71 drivers over 1 month, you are either bad at statistics, or your target level of safety is remarkably low. All you are doing with 71 people is making sure it's not so atrocious that you'll immediately have bad PR if you were to release it more widely. Hence the NDA.

As a simple rule of thumb FYI, you use the rule of three. Example:

If we have 71 drivers, and they each drive 2000 miles on City Streets Autosteer. None of them have accidents. That's 142,000 miles with no accidents.
You divide the 142,000 by three- 47,000. You now have a 95% confidence that your accident rate is 47,000 miles or more. But you have no idea what the actual rate is. It could be 47,001. It could be 17M. In fact, you can never really know what your accident rate is until you have some accidents. But you also don't need an accident if you can get enough data to go beyond your safety target.

I would argue that accidents don't matter at all. The key metric for determining whether it is safe enough for the general public to use it is safety disengagements.

The number of accidents is proportional to the number of safety disengagements times the reciprocal of how much attention the human safety drivers are paying. If people know that they're one of a carefully chosen group of people who have access, and that continued access is contingent upon them not screwing up, they're going to be a lot more careful than the general public, and so the accident rate should be low even if the technology is crap, at least up to the point where they get used to the system and become complacent, which they probably won't when you're talking about something that fails constantly.

When you open it up to the general public, however, most drivers are going to be much less careful, and the risk of accidents will be insane unless the number of safety disengagements is really low. So The fact that there have been no accidents yet, then, is mostly meaningless. They need to have almost no safety disengagements, or else it isn't ready to ship.

How many safety disengagements have you seen? My guess is that the rate in the previous beta is closer to one per 47 miles than to one per 47k miles. So that would make them even further away from their target.

 

[gearchruncher]

I don't think beta means what Tesla thinks it means. I'm pretty sure AP1 cars are still technically "beta"? I'm pretty sure that AP in my "FSD" car still has the beta label too...

The Model 3 manual uses the phrase "beta feature" 7 times (all for different functions), and the word "beta" 16 times. The silliest one being auto wipers.

The Chevy super cruise manual and Porsche Taycan manuals use it zero times.

 

[Bet TSLA]

Tesla tonight will begin the release process of the new version of FSD beta. They're starting by giving it to the same small sample of beta testers that they've been using. They have a baseline of performance for these people with their earlier FSD betas. They'll no doubt make sure that there are no regressions, i.e. that no new failure modes appear with these knows drivers. This will likely take a few tweaks, made about a week apart if history is any guide.

Depending on the number and severity of the problems, and how hard they are to fix, we'll likely see a wider release in a 4-8 weeks. Whether that will be "the button" or yet another pre-button interim step is anybody's guess. We still haven't gotten the release we were told we'd see by last Christmas for sure, so I'm not counting on anything.

 

[Bet TSLA]

For those of you debating Tesla's commitment to and priority of safety, please be aware that safety doesn't mean what you think it does. Elon has made it very clear that he thinks long term. If an approach can accelerate that advent of true FSD (i.e. no humans involved) by a year, then a few thousand deaths on the way is a fair price to pay. Today's level of carnage from human driving is truly grotesque. Safety first does *not* mean safety today, it means taking a path towards minimizing human suffering over the coming decades.

Be careful out there. And don't waste your time debating the permanently clueless.

 

[Daniel in SD]

For those of you debating Tesla's commitment to and priority of safety, please be aware that safety doesn't mean what you think it does. Elon has made it very clear that he thinks long term. If an approach can accelerate that advent of true FSD (i.e. no humans involved) by a year, then a few thousand deaths on the way is a fair price to pay. Today's level of carnage from human driving is truly grotesque. Safety first does *not* mean safety today, it means taking a path towards minimizing human suffering over the coming decades.

Be careful out there. And don't waste your time debating the permanently clueless.

Hey, I’m one of the clueless.
Clearly V8 had many unfixed bugs. Are you saying that releasing to a larger group before fixing known bugs would accelerate progress?
Also even if the sacrifice of thousands of lives would accelerate development it seems unlikely that society would accept that. Personally I’d like to see a convincing case made and not just take the word of a company that would profit from those deaths. To me it would be more convincing if the company making the claim only made money from those deaths after proving safety.

 

[gearchruncher]

For those of you debating Tesla's commitment to and priority of safety, please be aware that safety doesn't mean what you think it does. Elon has made it very clear that he thinks long term. If an approach can accelerate that advent of true FSD (i.e. no humans involved) by a year, then a few thousand deaths on the way is a fair price to pay. Today's level of carnage from human driving is truly grotesque. Safety first does *not* mean safety today, it means taking a path towards minimizing human suffering over the coming decades.

I am quoting this in whole for posterity because it is so insane.
This is peak, unabashed, Tesla kool aid. A defense of Tesla using indefensible arguments befitting the most evil people to ever exist.

You are arguing that a PRIVATE COMPANY has the right to KILL THOUSANDS OF INOCENT, UNINVOLVED AMERICANS with no oversight because it might save lives in the future if a single billionaire individual is correct, and this is a GOOD THING. All while under the cover of "beta" and "pay attention" and "Elon thinks long term unlike the permanently clueless."

I'm at a loss for words that you do not see the ethical issues with this. Please tell me you are not involved with medical care or anything else requiring any kind of ethics in any way. People fight wars over this kind of thing. History books have been written about this kind of thing. This is not OK.

 

[Dan D.]

The Model 3 manual uses the phrase "beta feature" 7 times (all for different functions), and the word "beta" 16 times. The silliest one being auto wipers.

The Chevy super cruise manual and Porsche Taycan manuals use it zero times.

Nowhere in the Model 3 manual do they define what they mean by the word "beta". It could mean what you think or it could mean what Tesla's lawyers think it means.

General convention is if you use a term you have to define the term. The place to do that is with an asterisk and footnote each time, or an entry somewhere like in the Disclaimers section. That they have deliberately not defined it in an otherwise comprehensive manual is suspect.

You cannot assume your audience knows what "beta" means. Does it mean "this feature is so unstable that we take no responsibility for it and you use it at total risk to yourself and others"?

By not defining it they have intentionally obscured whatever meaning it has.

 

[Bet TSLA]

Hey, I’m one of the clueless.
Clearly V8 had many unfixed bugs. Are you saying that releasing to a larger group before fixing known bugs would accelerate progress?
Also even if the sacrifice of thousands of lives would accelerate development it seems unlikely that society would accept that. Personally I’d like to see a convincing case made and not just take the word of a company that would profit from those deaths. To me it would be more convincing if the company making the claim only made money from those deaths after proving safety.

It's the permanently clueless who are best ignored. But no, I'm not saying anything about known bugs or how best to accelerate progress. I'm just saying that safety doesn't necessarily mean safety today. And I agree that unless Tesla stays within cultural limits of acceptable carnage (which are pretty extreme in the US), then it may cause problems for their engineering process.

 

[gearchruncher]

And I agree that unless Tesla stays within cultural limits of acceptable carnage (which are pretty extreme in the US), then it may cause problems for their engineering process.

You seem to think the USA is particularly bad in auto safety. We're at about 1:100M miles per fatality, while the best countries like Norway are 1:200M. A 2:1 ratio isn't really massive in safety, when even Elon is talking about a march of 9's, which are factors of 10.

This means that before Tesla is hitting our acceptable cultural level, they need to be better than 1:100M. Let's give it one more 9- 1:1B.

How long do you think it will be before Tesla can do a billion miles, hands off, without a fatality?

You say that "it may cause problems for their engineering process" as if you didn't suggest that an acceptable engineering process involves killing 3,000 people on your way to 1:1B mile autonomy. This is something no regulator has ever accepted. Do you think US culture would pay that price if given the choice? Would you sign up to be one of the chosen ones to sacrifice your life so that someone in the future may live?

 

[Matias]

Me too, but just in case it is worth repeating -- for the FSD Beta participants, be alert "there will be unknown issues, so please be paranoid."
View attachment 682947

https://twitter.com/x/status/1413306409693892613

I am glad he call it out again (I know he did a similar thing last October with the initial FSD Beta rollout).

”but there will be unknown issues, so please be paranoid. Safety is always top priority at Tesla.”

I think that’s self contradictory sentence.

 

[Matias]

If an approach can accelerate that advent of true FSD (i.e. no humans involved) by a year, then a few thousand deaths on the way is a fair price to pay.

I would not recommend this defence

 

[scubastevo80]

No button for me today - 2 cars with FSD and nothing. Really wish Elon would live up to his word…

 

[diplomat33]

No button for me today - 2 cars with FSD and nothing. Really wish Elon would live up to his word…

FSD Beta V9 was released today but only for a select few beta testers. So Elon did keep his word on this one. That's different from the button which Elon said is a month from now in the best case scenario. So we still need to wait for the button.

 

[scubastevo80]

FSD Beta V9 was released today but only for a select few beta testers. So Elon did keep his word on this one. That's different from the button which Elon said is a month from now in the best case scenario. So we still need to wait for the button.

I thought I saw a tweet saying he was expanding this? What’s the point of having only a few beta testers when they constantly say they need more data to make the software better. He even stated at one point we could email Tesla to be added to the program, which I did. This is unfortunately getting annoying.

 

[diplomat33]

I thought I saw a tweet saying he was expanding this? What’s the point of having only a few beta testers when they constantly say they need more data to make the software better. He even stated at one point we could email Tesla to be added to the program, which I did. This is unfortunately getting annoying.

The "email to be added to the program" part was cancelled AFAIK.