Yay! Nobody Cares about Tickle Me Elmo Musk

[Bartholomew Gallacher]

A little bit older, but still worth mentioning: Teslas can predict imminent crashes pretty well. If running on "autopilot", it will shut off itself about one second before the crash so that Tesla cannot be hold liable for it.

https://www.washingtonpost.com/technology/2022/06/15/tesla-autopilot-crashes/

 

[Innula Zenovka]

A little bit older, but still worth mentioning: Teslas can predict imminent crashes pretty well. If running on "autopilot", it will shut off itself about one second before the crash so that Tesla cannot be hold liable for it.

https://www.washingtonpost.com/technology/2022/06/15/tesla-autopilot-crashes/

I don't think that's what the article says, though.

According to the article,

Tesla vehicles running its Autopilot software have been involved in 273 reported crashes over roughly the past year, according to regulators, far more than previously known and providing concrete evidence regarding the real-world performance of its futuristic features.

and then, later on,

Some systems appear to disable in the moments leading up to a crash, potentially allowing companies to say they were not active at the time of the incident. NHTSA is already investigating 16 incidents involving Autopilot where Tesla vehicles slammed into parked emergency vehicles. On average in those incidents, NHTSA said: “Autopilot aborted vehicle control less than one second prior to the first impact.”

That means there's no indication, at least not in the article, that the autopilot shut down in the other 257 cases.

Also, the article qualifies the explanation of why the systems disabled themselves with "potentially." There's no indication that a court would actually regard the system disabling itself less than a second before the crash as exonatating Tesla in any way -- to my mind it would be surprising if a court were to find that way -- and it seems to me perfectly possible that the software disables itself not through any pre-planning but because it's crashed as a consequence of the overall software failure that caused the car crash.

 

[Noodles]

As much as I hate Musk, the other angle here is that, how many MORE time do human drivers get in accidents.

Its like when everyone was going on about the trolly problem for AI cars (which is dumb but I'll get there).

Even if an AI car kills a bus full of nuns every day, its still way less death than human drivers cause daily.

But the other side of this is that the Trolly Problem relies on "sudden failure". Which, for the nost part, does not happen, when proper maintenence is applied, which in a proper working AI system, will always be on. A human driver may not hear the break squal or will ignore it. An AI car can have a sensor and refuse to operate if the brakes are going bad in a zone that might result in "sidden failure."

It also relies on the concept of "human drivers really fudge traffic laws."

"What if you turn a blind corner and there are nuns and you can't stop in time, but you can instead run over a homeless looking guy????"

It doesn't matter, because the driver was going too fast around a blind corner. An AI car, will know, "I can't see around this corner, time to slow to 5mph so I can stop if there is suddenly nuns and hobos.". Its not going to drive recklessly like a human.

 

[Veritable Quandry]

Or more accurately, and I do not think we have near enough data at this point to answer the question, is how do the rates of accidents by humans driving a car directly or indirectly by software (because its humans all the way down) compare? Autonomous cars may have simply killed fewer people because there are fewer autonomous cars, and many accidents are ambiguous because systems like Tesla's require a human as a part of the system despite the marketing hype. Most of the fully autonomous tests are in urban areas with lower speeds. We don't know yet what will happen when they are regularly set loose on the freeways or in suburban areas.

However we do know that AI is writing mushroom foraging guides that are likely to kill people. Who knew Skynet would start as a gourmand?

 

[WolfEyes]

As much as I hate Musk, the other angle here is that, how many MORE time do human drivers get in accidents.

Its like when everyone was going on about the trolly problem for AI cars (which is dumb but I'll get there).

Even if an AI car kills a bus full of nuns every day, its still way less death than human drivers cause daily.

But the other side of this is that the Trolly Problem relies on "sudden failure". Which, for the nost part, does not happen, when proper maintenence is applied, which in a proper working AI system, will always be on. A human driver may not hear the break squal or will ignore it. An AI car can have a sensor and refuse to operate if the brakes are going bad in a zone that might result in "sidden failure."

It also relies on the concept of "human drivers really fudge traffic laws."

"What if you turn a blind corner and there are nuns and you can't stop in time, but you can instead run over a homeless looking guy????"

It doesn't matter, because the driver was going too fast around a blind corner. An AI car, will know, "I can't see around this corner, time to slow to 5mph so I can stop if there is suddenly nuns and hobos.". Its not going to drive recklessly like a human.

What will happen when the damn thing shuts down on the way to the emergency room? What will happen when an AI ambulance shuts down due to equipment failure before it reaches the emergency room?



No thank you.

 

[Argent Stonecutter]

An AI car, will know, "I can't see around this corner, time to slow to 5mph so I can stop if there is suddenly nuns and hobos.". Its not going to drive recklessly like a human.

I used to argue this, but Musk proved me wrong. I think it might have been removed, but at one point Tesla had an "aggressive mode" that would do things like California stops.

 

[Noodles]

I used to argue this, but Musk proved me wrong. I think it might have been removed, but at one point Tesla had an "aggressive mode" that would do things like California stops.

That feels like it puts the liability on the car manufacturer then.

It doesn't surprise me given Musk doesn't seem like the type to give a shit about the law.

 

[Veritable Quandry]

Our Self-Driving Cars Will Save Countless Lives, But They Will Kill Some of You First

I’m going to say something you wouldn’t expect from the CEO of a car company: America’s roads are a warzone. Over 40,000 traffic fatalities in 2021...

www.mcsweeneys.net

 

[Cindy Claveau]

 

[Innula Zenovka]

Musk Secretly Used Starlink to Foil Ukrainian Drone Attack on Russian Ships: Report

The billlionaire allegedly feared the strike on Crimea could trigger nuclear war.

www.thedailybeast.com

 

[Essence Lumin]

Musk Secretly Used Starlink to Foil Ukrainian Drone Attack on Russian Ships: Report

The billlionaire allegedly feared the strike on Crimea could trigger nuclear war.

www.thedailybeast.com

That article is an excerpt from a cnn article which is here.

https://edition.cnn.com/2023/09/07/politics/elon-musk-biography-walter-isaacson-ukraine-starlink/index.html

 

[Free]

One might need to point out that Starlink's services are paid for. Apparently Elon is not a capitalist.

 

[WolfEyes]

How unpatriotic and un-American of Musk.

 

[detrius]

How unpatriotic and un-American of Musk.

He's originally from South Africa, acquired Canadian citizenship through his mother and became a US citizen for convenience.

I'm pretty sure he thinks the way loyalty works is that countries should be loyal to him and not expect it to be the other way around.

 

[Innula Zenovka]

Elon Musk and the west’s fascist fifth column

The weekend political column is free to read.

nickcohen.substack.com

Let’s take a pause here. A US citizen acted against the interests of Ukraine, a US ally, and intervened to prevent what might have been a militarily decisive attack on a US enemy’s navy. How incredible that a private individual has such power and can exercise it without any constraint from the democratically elected US government.

I would say that Musk is a contemporary version of Henry Ford, who supported Hitler and Nazi ideology in the 1930s, as Musk supports Putin today. But Henry Ford was a nobody in comparison to Elon Musk. Ford only made cars and trucks, and had no power to intervene in US foreign policy beyond the funding he gave to fascist propaganda. Musk directly intervened in a European conflict on the Russian side. He did so after talking to an unnamed Russian government official who, Musk claimed, encouraged him to believe that the proposed Ukrainian drone attack could escalate into a nuclear conflict.

In old fashioned language Musk was treasonably aiding the enemy in time of war. If such denunciations are too much for delicate 21st century sensibilities to take, can we at least say that through his control of Starlink, and of Twitter (now X) Musk is a Kremlin agent of influence? As if to prove the point, the EU showed, that since taking over Twitter, Musk allowed Kremlin backed accounts to expand their “reach”. Musk himself has propagated every variety of conspiracy theory, and was last seen blaming his financial problems on a conspiracy of – oh, go on, have a guess – of Jews! (Who could have seen that one coming?)

 

[WolfEyes]

He's originally from South Africa, acquired Canadian citizenship through his mother and became a US citizen for convenience.

I'm pretty sure he thinks the way loyalty works is that countries should be loyal to him and not expect it to be the other way around.

Funny. I recently was talking about how Musk was born in South Africa on this very forum. Something about why he hasn't been deported along with all the blacks like conservatives have liked to threaten for decades.

While my post was made tongue in cheek, it was meant to be humorous.

 

[Innula Zenovka]

Reassuring

 

[Katheryne Helendale]

Reassuring

 

[Kamilah Hauptmann]

Reassuring

 

[Noodles]

Terminal moneys?

Also, in my one college psychology class elective, the professor drilled in that "Chimps are not monkeys", because chimps are more closely related to humans than monkeys. Seems like chimpanzees would be a better choice for testing this thing than monkeys.