Tesla Autopilot 'Predicts' Accident Before It Happens (engadget.com) 186
A dash cam footage suggests a Tesla on Autopilot may have predicted a nearby freeway crash before it actually happened. A video showed that a Tesla car driving on a highway in the Netherlands started to beep a few seconds ahead of two cars colliding with each other in front of it. A Tesla representative confirmed to media that the beeping heard in the video is indeed the sound of Autopilot's Forward Collision Warning. Elon Musk tweeted a news article about the incident, adding more credibility to the matter. From a report on Engadget:Tesla's Autopilot 8.0 has a particularly clever feature: it uses radar to track road activity two cars ahead, helping it avoid danger that you wouldn't normally see. And it now appears that this tech just averted a disaster. Dutch Model X owner Frank van Hoesel has dashcam footage showing his electric crossover reacting to a bad highway crash before it even starts. As you can hear in the video, the Model X's Forward Collision Warning system starts braking when it detects the SUV two vehicles ahead coming to an abrupt stop, even though the driver of the car directly behind it is unaware. The result? Van Hoesel's EV remained untouched when it could easily have contributed to a pile-up.
cars with radars (Score:5, Funny)
Re: (Score:2)
next i want to hear the alert for missile lock and see it auto deploy counter measures..
You'll probably have to wait for their middle east rollout
Quick Pick (Score:2)
Wake me up when my Tesla can predict tonight's Lotto numbers.
Re: (Score:2)
It can. DE AD BE EF and the powerball is 69
Even better than mitigating a developing accident (Score:2)
would be for the small car in front to be piloted by a non-human.
Predicts? (Score:2, Insightful)
It can see the cars ahead and it saw the SUV breaking so it reacted. It did not predict anything other than a vehicle in front is breaking and it started breaking faster than a human can react.
Tomorrow on slashdot..... "Scientists perform witchcraft by telling us that we will have a solar eclipse this next summer, and your jaw will drop and be speechless with a blown mind"
Re: (Score:1)
it saw the SUV breaking so it reacted.
The SUV didn't break until it rolled over. But it was braking until it broke.
Cutting through contention and nonsense (Score:5, Informative)
Gee, the comments on this so far stink. People seem to be prejudiced.
First, the dashcam video presented is not from Tesla, it's from an independent car fan who seems to have gotten it from the driver. It was then retweeted by Elon Musk.
The activity demonstrated is the Tesla Autopilot triggering on sudden deceleration of the second car ahead, which is not clearly visible from the dashcam view (and presumably equally not clearly visible to the driver). It appears that autopilot warns and brakes.
Autopilot does not predict what happens to the cars ahead (although in second-car detections it's probably implied) and "anticipate" would be a better word anyway. Autopilot anticipates that the Tesla will hit something if it doesn't brake.
Soneone on twitter pointed out that he heard the "disengage" sound after the warning sound, indicating that the driver brakes. I don't hear if, but I don't know what it sounds like. Does anyone else hear it?
Someone pointed out that the Joshua Brown accident might not have happened if the radar had worked then. Yes, it might not have. One should also point out that Brown was speeding and apparently not looking. Despite its name, "autopilot" is not ready for the driver to disengage.
Re: (Score:2)
It looks to me like the poor quality of the video makes it harder to see the second car ahead than it would be in real life. The car which suddenly stopped looks like it's a minivan, and even with the low resolution of the video, at the moment the warning system beeps you can c
Re: (Score:3)
I don't really trust myself to catch that sort of situation in time to brake 100% of the time. I'd be happy to have a system that could help me that way.
Next question: was the Tesla maintaining a sufficient stopping distance? I tend to give more space than that.
Re: (Score:2)
You can select the following distance between 1-7 by twisting the cruise control lever. It's a time based system, so it will lengthen as speeds increase.
Re: (Score:2)
Predicts accident (Score:1)
Re: (Score:2)
The notion of prediction is not applicable to a partially observable, stochastic environment.
Of course it is! A weather forecast is a prediction. Someone saying that a particular horse is most likely to win a race is a prediction. Of course in all cases there is only a certain probability of it being correct, but it is still a prediction.
It's neat, but also a clever marketing campaign (Score:2)
And not just for Tesla either. Interesting how the video is published through the Twitter account of a personal injury consultant, no? It's not linked from the Tweet that's mentioned in the article and news outlets are actually conversing with this guy for the rights, even though their involvement seems lateral at best. Look for the video online and you find the same dude posting the tweet wherever they will have it (instead of just posting the YouTube copy that's also available, of course https://www.youtu [youtube.com]
Good driving (Score:2)
Save lives (Score:5, Insightful)
This will save lives, full stop. You can crap all over Elon if you want, but the guy actually gets shit done. No it's not perfect, but it keeps getting better and it is at the point where it will save lives.
Re: (Score:2)
Predict? (Score:2)
More like... (Score:2)
Re: (Score:1)
Must have been those dangerous radar waves it uses. You know the military uses those in advanced weapon systems?
And it's the same type of thing a microwave uses to boil water. So why is tesla shooting them at innocent people, huh?
Re: (Score:2)
You need to be raped by a wild boar and left to die, alone, in the woods.
Re: (Score:2)
If you look at Dog-Cow's comment, there is a Parent link at the bottom of it. This button will highlight the post to which he was replying, and let you know that this is the comment he believes caused the AC to diserve such wonderful treatment.
Because you have a brain the size of a pea.
Do you live in some backwater redneck county that doesn't provide basic science education?
Go put on your Tinfoil hat.
Re:No. It didn't "predict" anything. (Score:5, Insightful)
"Predict" is exactly what it did. "Predict" doesn't mean the mystical fortune-telling you seem to think it means.
Re:No. It didn't "predict" anything. (Score:5, Insightful)
It predicted that *it* would crash if it didn't slow down. It didn't predict jack about the collision between the other cars.
Re:No. It didn't "predict" anything. (Score:4, Informative)
Of course it didn't predict the actual accident, but it did see the car two cars ahead abruptly slowing down (by bouncing the radar signal underneath the car in front) and reacted by braking in time. The car directly in front of the Tesla never even touched the brakes.
That's pretty impressive, and the model S and X are the only cars with that capability.
Re: (Score:1)
Except sometimes you're driving behind a large van and can't see anything ahead of it. The Tesla radar actually bounces its beam underneath the car in front of it.
Re: (Score:1)
If the car in front of you hits a stationary car, you'll need to have left quite a lot of distance (well more than the recommended 2 seconds) to come to a stop in time. Nothing to do with tailgating, just physics. Recommended distances are based on reaction time, assuming the car in front is braking like you do. If it comes to an abrupt stop due to a collision, those two seconds are not going to save you.
So that's why good drivers try to look ahead (like I do), but sometimes you just can't. And the the Tesl
Re: (Score:2)
recommended 2 seconds
At 25 miles per hour, maybe. It's at least 4 seconds at highway speed - reaction time is too critical. You can't only prepare for tailgating, you really need to assume that anyone ahead of you can crash at any moment.
Re: (Score:2)
Too much. Too many close calls because of stupid people.
Re: (Score:2)
Ever had a van cut in front of you on the freeway? There's some time before I can establish a safe interval, speaking for myself.
Re: (Score:1)
Well, they do advertise it as bouncing underneath the car in front, so it should also work behind a van without a rear window.
Re: (Score:2)
It did literally "predict" the accident, just like I can "predict" that my pen will hit the floor should I drop it. It responded to inputs and reacted. In this case, it responded to inputs, drew a conclusion, and responded in a fashion that a human would have likely missed. There's definitely something there.
Still, I think an equally appropriate title would be "Tesla Autopilot avoids getting into an accident." Every action the autopilot takes (or even actions taken by human drivers) are based on predictions
Re: (Score:2)
It did literally "predict" the accident, just like I can "predict" that my pen will hit the floor should I drop it.
This presumes that the "accident" is only the collision and not the entire stream of events. The stream of events was not predicted, just the outcome after initiating events created certainty, or near certainty of the outcome. It seems that the accident was detected and avoided rather than predicted in this sense
Or could similar conditions result in a prediction of a collision, but not result in one? Your laptop drop has certainty whereas the events leading up to the collision may have had uncertainty whi
Re: (Score:3, Informative)
They updated the radar software. Instead of using a single beam, it now scans around and tracks multiple objects. Uses a constantly updated database to avoid false alarms for street signs and such. Free over-the-air update for all autopilot Teslas. It would now see that truck and react accordingly.
Re: (Score:2)
Well, OBVIOUSLY the car that slammed into the stopped car missed that.
Re: (Score:2)
Re: (Score:2)
By the time it "predicted", the accident sequence was already well under way, starting with the sudden deceleration of the larger vehicle. A 100% probability of loss-of-separation is an easy enough "prediction" so as to be called a calculation.
Re: (Score:2)
There is such a thing as predictive modeling. Maybe the word "predict" bothers you somehow, yet the computer may have understood there was likely to be a crash before it happened. Maybe it only understood that it was a good time to brake. Giving probability to a future possibility sounds like prediction to me.
Re: (Score:3)
Re: (Score:3, Insightful)
But - but - but - Elon Musk tweeted a news article about the incident, adding more credibility to the matter!
Re: (Score:2)
Re: (Score:2)
Thanks. I'm really getting tired of the tech-billionaire-as-culture-hero mentality in the stories here.
Re: (Score:2)
Re: (Score:2)
But - but - but - Elon Musk tweeted a news article about the incident, adding more credibility to the matter!
That behavior is predictable.
Re: (Score:2)
Would a person notice and react in the same way? If they are even semi-competent to be driving, yes. If they belong to either the 'stare directly at the car in front of you at all times' or 'my phone is so much more important than driving' camps, then no.
Re: (Score:2)
If they are even semi-competent to be driving, yes. If they belong to either the 'stare directly at the car in front of you at all times' or 'my phone is so much more important than driving' camps, then no
So IOW, it's superior to the average driver? Counties with decent driver's ed might well mock, but others ought to consider adopting the tech en masse today.
Re: (Score:2)
The Netherlands, where this accident happened, have very strict driver education laws. The exams are both theoretical and practical. Most people need about 40 hours of driving lessons to even qualify for the exam. A lot of people don't even get it on the first attempt and have to re-take the exam later.
Even then you still see a lot of folks driving around who should never have gotten their license, or should have returned it. Tesla's and other cars with this software should be a major improvement in road sa
Re: (Score:3, Informative)
The car directly in front did not even touch the brakes. It was a very sudden and unexpected deceleration two cars ahead.
Sure, I try to keep track of multiple cars in front, driving at the far left of the left lane so I can see any brake lights come on ahead, but sometimes you just can't (behind a large van, for example) and noticing such a sudden stop while the car directly in front of you keeps going... no, most people would just have crashed into it. Hell, most people can't even stop in time if the car d
Re: (Score:2)
You're correct, of course, that there was no mystical prognostication powers at work here. But in the same way we can predict a spacecraft's flight through our solar system based on a known starting state, so too can we say Telsa "predicted" the accident when it was apparent that the laws of physics were in control instead of the drivers.
Prediction doesn't necessarily imply the existence of magic. After all, meteorologists predict (although they prefer "forecast") future weather a week from the current da
Re: (Score:2)
After the semicolon, you begin to explain (at a high level, without much detail) how it made the prediction.
And by teaching people how to predict things (observe a system's current state and extrapolate where it's going) you are helping to teach people that computers are not magic. Good for you. I didn't even know that anyone was trying to prese
Re: (Score:2)
And the car directly behind was not unaware- he attempted to evade, but didn't quite make it, which is why the rear-ending hit the corner of the SUV and the red car only had damage on the driver's side
Re:Confirmation bias? (Score:5, Informative)
So we live in an age where the only acceptable stories that can be reported in the media are negative ones?
Re: (Score:2)
So we live in an age where the only acceptable stories that can be reported in the media are negative ones?
No, that's not true at all....
It doesn't matter if it's positive or negative.
Re: (Score:1)
Exactly, the Tesla wasn't rear ended. Either was an other Tesla behind it or a mere human that did the same as the Tesla infront of it or the human kept an acceptable distance instead of tailgating.
Re: (Score:3)
Exactly, the Tesla wasn't rear ended. Either was an other Tesla behind it or a mere human that did the same as the Tesla infront of it or the human kept an acceptable distance instead of tailgating.
Or the Tesla autopilot gauged the distance to the trailing car and moderated its braking to ensure that the human behind had sufficient time to slow. That's exactly what I'd do in that position; any experienced heavy-traffic freeway driver knows that the first thing you do when you start braking to avoid something in front of you is check the rearview mirror so you can adjust your braking appropriately to avoid getting yourself rear-ended -- or to decide if you need to swerve into the emergency lane. Of cou
Re: (Score:2)
Fun fact: a lot of accidents with moderate damage occur where the driver presses the brakes deeply, then releases because he or she realizes someone might be behind them, and THEN collides with the car in front. Mercedes specifically created an emergency brake system to counter this. It would keep applying the brakes when pressed "vigorously" for a few seconds after the driver released them. Gives you a hell of a surprise the first time you feel it...
Re: (Score:2)
Fun fact: a lot of accidents with moderate damage occur where the driver presses the brakes deeply, then releases because he or she realizes someone might be behind them, and THEN collides with the car in front. Mercedes specifically created an emergency brake system to counter this. It would keep applying the brakes when pressed "vigorously" for a few seconds after the driver released them. Gives you a hell of a surprise the first time you feel it...
Ouch. That seems like a cure worse than the disease.
Re: (Score:2)
The car behind the Tesla could see the Tesla's brake lights. The car in front of the Tesla never braked.
Re: (Score:2)
If the driver behind the Tesla was any good, they were looking well beyond the collision and reacted before seeing any brake lights on the Tesla.
Re: (Score:2)
If you watched the video, you'd know that it wouldn't have helped in this case. That the vehicle in front stopped is not visible until after the car between it and the Tesla crashed into it.
Re:Confirmation bias? (Score:4, Insightful)
Re:Confirmation bias? (Score:5, Informative)
Did Tesla also report any/all instances where the forward collision warning sounded, regardless of whether or not a crash subsequently occurred? Otherwise this is just PR.
First off, an erratic driver is obviously worth paying attention to, so it's worth having the beep for near-crashes as well.
Also, in terms of the warning system's efficacy effects this probably isn't relevant unless it beeps so much as to cause drivers to ignore it. The false positive rate could be 75% and it still probably wouldn't beep more than once a week at the most (depending on where you live / Boston joke goes here.)
Re: (Score:3)
Did Tesla also report any/all instances where the forward collision warning sounded, regardless of whether or not a crash subsequently occurred? Otherwise this is just PR.
Tesla never reported any instances where the forward collision warning sounded. It's hard for it to be PR when it's not Tesla doing the reporting.
Re: (Score:3, Insightful)
You mean the ones that had the advantage of having both cars in view, instead of a blocked view like the Tesla (driver) had. Ah, those cars.
Bert
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
There are fantastic-seeming things that are plausible in our lifetimes. Like, say, general AI or the scaling up of countless nanotech technologies that are still in their infancy. It's a waste of street cred for him to talk about this stuff. It's not that I fear his failure; it's that I fear young geeks becoming disillusioned when they realize his failures were mostly due to him being ful
Re: (Score:2)
Getting lots of mass there cheaply lets the other solutions be a lot less fancy, though.
Re: (Score:2)
Your hyperloop "flaws" are absurd (Score:3)
This requires some basic independent critical thinking
Wow, talk about a real "physician heal thyself moment".
really isn't all that fast (only 2x Maglev),
A 2x speed improvement is a huge deal for any traveller.
won't be cheap (a jet engine, linear motor AND an airtight tube instead of a track? And how many ticket-paying people can cram into these cars vs. traditional rail?)
It may not be cheap but it uses a lot less energy per trip, and requires less maintenance, AND has dramatically safer failure mode (i.e.
Re: (Score:2)
A 2x speed improvement is a huge deal for any traveller.
To be partially belied by the increased security concerns. A tube rupture isn't like a break in the track. The wall of air that goes rushing on both directions at hundreds of miles per hour poses a huge hazard, and the proximity to the interstate makes this worse in several ways. There *will* be security to get onto this thing.
Just the fact you don't have to go through security means an automatic two-hour total reduction in travel time before you factor in actual trip duration
The duration of the proposed route was only what, something like 30 minutes? (I'm not looking it up right now, but I know it wasn't long.) It doesn't take much of a security delay fo
Re:Your hyperloop "flaws" are absurd (Score:4, Informative)
The wall of air that goes rushing on both directions at hundreds of miles per hour
They Hyperloop uses a near vacuum so... yeah. And even if it didn't about ten feet away it would be perfectly safe in a rupture. It's crazy to think this is any danger except for the moron that tries to pierce the tube.
If what you were afraid of was an issue every person with an air compressor would be in mortal danger.
You realize the joints between the tubes probably can't be a plain seal made of the same material as the rest of the pipe, right?
You realize the tube replaces potentially hundreds of airplanes, right? And that it's easy to reach the tube to maintain... and that maintenance and inspection can be easily and highly automated...
Discounting the increased security required for what is going to be a hugely attractive target for terrorists.
*cough*bullshit*cough. It is simply idiotic to think this is anywhere near the target planes are given that so few people would be affected compared to an airline disaster. Again, if you get a plane you get to kill the people on the plane PLUS any ground target you like. Again, if you take over a tube you MAYBE can kill some people on the tube before they stop you? And you can't re-aim the tube to attack anything else. I do not understand why you fail to grasp the magnitude of difference in attack vectors here, truly mystifying.
everyone on that car is going to die.
Nope, it would simply come to a stop. It might be rough but there's no way a single person is even seriously hurt by a breach in the tube. Just stuck for a bit. There are about ten billion ways you can devise the car to have a failsafe so that loss of pressure simply means the car stops.
Again, a mere 15 minute delay due to security removes most of that advantage
15 >= 120... Nope, does not compute. Not to mention since hyperloops are run at regular intervals it's not nearly as big a deal if you miss one, you just get the next - so there's no need ot be as cautious in traveling. I tend to show up 2-3 hours early for a plane - and I have pre-check.
The hyperloop is a watering-down of a very old idea that was impractical ~100 years ago
I wonder if material science has improved in the last 100 years. NAH CANT BE.
I'm going to let you have the last word on this since you seem to be way more unreasonably fanatically against anything Musk might have thought up than anyone who is for Musk's ideas ever was.... in the end a fight with a zealot is simply a waste of my time as you cannot learn and will not change.
Re: (Score:2)
They Hyperloop uses a near vacuum so... yeah. And even if it didn't about ten feet away it would be perfectly safe in a rupture. It's crazy to think this is any danger except for the moron that tries to pierce the tube. If what you were afraid of was an issue every person with an air compressor would be in mortal danger.
Air is pushed (accelerated) into the tube by the weight of the air around it (air pressure.) That weight (yes, literal weight, caused by gravity) keeps "falling" down the tube, picking up speed as it goes, and since it's a near-vacuum there is very little to slow it down. Acceleration over many seconds gives you a significant speed. I'm unfamiliar with the behavior of compressible fluids to know whether or not this would actually reach the speed of sound, but given what I do know it sounds plausible enough
Re: (Score:2)
Pseudoscience (Score:2)
If there's something wrong with the assertion that the air will accelerate up the tube (not travel c
Re: (Score:2)
Re: (Score:2)
The force moving the air in the tube is constant. It is the pressure of the atmosphere multiplied by the cross section of the breach. So, as more air enters the tube, the total mass in the tube goes up, reducing the overall acceleration. Also, near vacuum is not vacuum, and there will be friction between the air and the interior surface of the tube, reducing the effects further. I'm also ignoring turbulence - which is probably the biggest factor. At some point, acceleration will get down to zero, althou
Re: (Score:2)
The "vacuum is not a vacuum" thing has been played to death and is bullshit. 1/1000 atmosphere is close enough to ignore whatever is in there. This is akin to taking off your sunglasses before stepping on the scale.
friction between the air and the interior surface of the tube
Pretty darn sure this is irrelevant as well. If you cannot ignore these minor details...
So, as more air enters the tube, the total mass in the tube goes up, reducing the overall acceleration.
Bah. Show your work. The mass in the tube is already moving quickly in one direction and losing negligible speed. That mass
Re: (Score:2)
https://www.reddit.com/r/AskPh... [reddit.com] https://www.reddit.com/r/AskEn... [reddit.com]
Re: (Score:2)
pressure differential propagates backward once air is in the pipe
Makes no sense. The air is already moving at a good clip this way =====> in an effective vacuum. There's nothing I can see to slow it down quickly enough to do that. The air pressure resistance just inside the breach should be negligible until the entire tube is filled.
you could add air between a breech and the train to provide a barrier
I didn't say it was impossible to do. I said it was going to be expensive, on top of an already very expensive project.
The other important thing is that air, unlike a solid can flow around things.
If it's moving at near the speed of sound as Thunderf00t claims (and in the absence of a good ex
Re: (Score:2)
And after I did a tiny bit of research last night, it appears that air is really not all that compressible even at those speeds. I fail to see how that wouldn't result in, at the very least,
Re: (Score:2)
Re: (Score:1)
Re: (Score:2)
If you're interested in an actual answer here (as opposed to mere smugness or hero worship), I've just taken up the matter at
Re: (Score:2)
air can travel around the train thus reducing pressure differential
Not something you can hand wave away, by the way. You're talking about aerodynamic displacement around and behind the vehicle generated from forward momentum. We can't assume this displacement will just automatically happen at a fast enough rate to ensure that the vehicle doesn't fatally decelerate.
It's seeming as though things are a bit more complex than I initially thought, but that doesn't give you any points for your hand-waving dismissals. In particular, anyone who flatly says that the air wouldn't
Re: (Score:2)
Re:Horrific way of reporting it (Score:5, Informative)
So does that mean it also "predicted" that it would kill that driver who slammed into a transport at high speed?
That was before the 8.0 update. Unlike other cars, Teslas get updates. The old software just used a single radar beam, which passed underneath the high trailer in the accident you are refering to. The new software lets the radar beam scan around to create a point cloud, tracking multiple objects. In an identical scenario, it would have seen the trailer and braked in time. And it can also look two cars ahead by bouncing the radar signal on the asphalt, which is what saved it in this latest video.
Re:Horrific way of reporting it (Score:5, Informative)
Well it worked great that one time out of a few hundred.
Tesla rarely has customer dashcam videos for every time it "has" worked...we mostly only get to hear about the failures, which given the number of Teslas on the road that used autopilot in 2016, are considerably few.
...airbags and seatbelts have saved countless but we don't see an article praising how well it works.
You mean we don't see them anymore...there are literally thousands of articles and advertisements about airbags and seat belts going over the past 60 years.
Re: (Score:3, Insightful)
Re:The Tesla is psychic - it knows when all accide (Score:4, Insightful)
You're right. This wasn't a "prediction". The car's radar is able to see two cars ahead and saw that traffic was stopped so it put the brakes on and alerted the driver. Impressive performance but not psychic.
Re: (Score:2)
Not even particularly impressive. The autopilot was allowing plenty of following distance, even a human could have reacted in that timeframe. I could see the back corner of the black SUV stop the same second in the video that the beeping started. I would have reacted in exactly the same way.
Re: (Score:2)
Accurately maybe. Strictly? I'm strict about it to the point of practicing the martial art of antitrafic (when in a jam, slow down to 5mph below average speed, to build a space in front of you and avoid stop-and-go starts). I wish more people knew about it, there would be far fewer accidents and we'd all get where we need to go faster.
Re:The Tesla is psychic - it knows when all accide (Score:5, Insightful)
I have seen accidents about to happen and avoided them many times in my many years driving, as I'm sure most of us have. In none of those cases did we go around saying we 'predicted' those accidents.
This may say more about your use of english language than anything else.
You saw something about to happen did you?
predict:
verb
say or estimate that (a specified thing) will happen in the future or will be a consequence of something.
English is a rich language. Why use a word like "detected" when you are describing an event that hasn't happened yet? If you're talking about traffic abnormalities then yes you "detected" it. If you're talking about the impending collision then you "predicted" it and then "avoided" it.
But my guess is you were more concerned with heaping shit on Musk / Tesla, which is a shame because all you showed is that his language skills are more impressive than yours may actually be.
Re: (Score:2)
STOP! I predict we will have an accident (Score:2)
Slams on brakes. Car behind rear ends. Real prediction, and correct.
[Get Smart, with a psychic that KAOS wants to kidnap.]
Re: (Score:2)
I bet you wouldn't use that term if you had just avoided an accident.
That depends entirely if I "predicted" the accident before it happened or simply "avoided" it once two cars hit themselves. That's how english works. We don't adjust the language to suit context. The language itself is clear.
Re: (Score:2)
The language itself is clear.
No, the english lanquage can convey different meanings with the same words, and choice of words in a particular context can convey different meanings or impressions. "predict" sounds a lot more impressive that "detect". If you saw two cars obviously about to collide and reacted accordingly, you would not go around telling people you predicted the accident.
Re: (Score:2)
and choice of words in a particular context can convey different meanings or impressions
Welcome to my point.
"predict" sounds a lot more impressive that "detect".
Indeed because the act of predicting is more impressive than the act of detecting. Now if you want to take issue with the specifics of the instance and call Musk a liar then be my guest. But what you are claiming is that people don't use words like "predict" to describe accidents which either means they didn't "predict" them or they don't have a sufficient grasp of the English language to make their point correctly.
If you saw two cars obviously about to collide and reacted accordingly, you would not go around telling people you predicted the accident.
I would, because I'm not an idiot who dumbs down language to borderline n
Re:The Tesla is psychic - it knows when all accide (Score:4, Insightful)
Musk picks terms to make things sound more impressive than they may actually be. I have seen accidents about to happen and avoided them many times in my many years driving, as I'm sure most of us have. In none of those cases did we go around saying we 'predicted' those accidents. We say we 'detected and avoided' them. .
Yes but it detected the braking of the car ahead of the car in front, which didn't seem to brake at all until the crash. From the video the braking car was barely visible. That's pretty impressive.