[CptFribble]

Cars are not software, and it's ethically and morally wrong to test them in production.

Musk is playing games he doesn't understand, and wagering other people's lives to do it. He should at minimum understand that he has many, many fans who trust that he knows what he's doing, and will not expect Tesla's products to be cutting corners on safety and testing because "simpler is better" and "move fast and break things."

There are ways to find optimizations without removing a bunch of stuff and just shipping it like that to the general public: it's called engineering.

If Musk really wants to find ways to optimize the concept of a car further, he'll have to give up on point #4 and accept that it's going to take a lot of test cycles to figure out what works and what doesn't. Rushing out half-baked concepts that are likely missing key safety features because "let's see what happens" is exactly the kind of braindead approach to engineering management that is keeping me approximately 10,000 miles from anything Musk is in charge of.

[generalizations]

#4 is the feedback loop that enables rapid improvement. Long as the experiments aren't life-threatening, are you really proposing that they only test new ideas in fake limited artificial environments, rather than in real-world environments where they can encounter the full scope of possible failure modes?

[RaftPeople]

> Long as the experiments aren't life-threatening, are you really proposing that they only test new ideas in fake limited artificial environments, rather than in real-world environments where they can encounter the full scope of possible failure modes?

Are people good at reliably determining which experiments are life-threatening? Especially when those people are under pressure to move fast?

[eropple]

When it comes to multi-ton death machines, selling something you don't know is safe is the same thing as selling something you know isn't safe.

[aeternum]

No one knows what is safe, those that purport to know exactly what is safe are the most dangerous.

Safety is the new snakeoil. Add just a drop and a sprinkle of "think of the children" and you can sell your BS to anyone.

[yndoendo]

Number of European countries require safety to be engineered into the product. Example, where I work, a machine in automation had a risk assessment that it produced 1600 newtons of clamping force, same biting force of an adult panda, could take off limbs. This machine could be sold in the USA and not the Europe. Re-engineered to be safe and sold in Europe, machine cannot even take off a finger.

USA is poorly regulated to keep operators safe. Designing for European mark means you can market safer than the competition in USA and sell in Europe at the same time. This also remove the need to build in safety guards or use light curtains.

Which one would you or your family members like to use day-in day-out, the 1600 newton limb remover built on USA standards or the one built around safety for Europe?

[aeternum]

Depends, I might opt for the limb remover if it is simpler and works more reliably. Time is money and I'm confident I could use either properly so I'd likely waste more of my limited lifespan fixing the safe but more complex and more unreliable machine.

If there's no cost to the safety factor then sure, put it in. However there's almost always a cost. We all pay with a portion of our life every time we stand in an airport security line and every time we click accept to one of those EU cookie popups in a futile attempt to keep our data safe. With each click, billions of humans have a precious few moments severed from their short time on earth, moments that could be spent with loved ones.

[eropple]

I've been around a lot of people who say that sort of thing and it's why in my neck of the woods we note that "I never get distracted, I'm never tired, I always work perfectly" is the mantra of a woodworker who has eight fingers. (He didn't learn the first time.)

Time is money, but you can't get a finger back. And neither can your employees who work for you, who the EU regulations are, rightly, more concerned with than your bottom line.

[7952]

It is perfectly fine to sell a dangerous product as long as the people exposed to it have consented in an informed way.

[brokenmachine]

I would hereby like to revoke consent for Tesla fans "Fully Self Driving" their deathtraps on public roads that I use.

[aeternum]

You are probably a worse driver than Tesla FSD, I'd like to revoke consent for you to use the public road.

[eropple]

Yeah! How could you know anything really, you know? It's just impossible to know that you should securely affix, and test the security of, the cover of an accelerator pedal on a motor vehicle. There's no prior art to accelerator pedals and no examples of what happens when you make a bad one.

Or, you know, the other thing.

[redavni]

So why do we sell cars at all? They are the 3rd leading cause of death since before Elon Musk and yourself were born.

[coldbrewed]

There's a difference between selling a multi-ton vehicle that has crumple zones and curved lines and selling a multi-ton vehicle that is designed to tenderize pedestrian rib cages. This comment also applies to today's pickups and SUVs; but while those vehicles are pretty nightmarish for the safety pedestrians and other drivers the CT is a further escalation of matters through both design and build quality.

[eropple]

I don't understand the motivation behind such a question. We sell cars because the benefits to motor transportation outweigh the safety risks in aggregate. That aggregate tradeoff is nonresponsive to the singular case of YOLOing avoidable risks on a Tesla.