Hacker News Reader

Comment by lutusp | Hacker News Reader

> I can't shake the feeling that one cannot take anything uttered from Elon Musk's mouth at face value when it concerns his baby ...

There's no need for that -- Musk has the car's logs, which describe the miles driven, the speeds, and the battery's state of charge. Under the circumstances, this is not a case of "he said, she said" -- i's a matter of unambiguous computer records.

If the computer records contradict what Musk said, he's in big trouble, but I doubt it will come out that way. And consider the consequences if Musk reported something different than the records show -- and Musk knows this. Why would he knowingly sabotage his own company?

View on HN

So why hasn't Musk or Tesla made an official response yet?

If Musk really did reach out to Broder on Friday, then Tesla should have had a skeleton press release ready to go on Monday, pointing out the major mistakes that Broder made, and pointing out that they're planning on getting out more supercharging stations, so this experience won't be repeated. It should have been a neutral response to the criticism, and it would have highlighted the infrastructure work that Tesla is doing.

Instead, Musk has claimed that the Times' account was "fake", which is a huge accusation to make, because that's a personal attack on Broder and the Times, and, let's face it, the NYT is probably much better at CYA and public perception management than Tesla or Musk are.

"Fake" was probably a poor choice of words, and typical of the knee jerk commentary you see on social media.

That doesn't excuse it, but you'll notice in his subsequent interviews that the language he uses is much more tempered, probably because he got a few nervous phone calls from his legal team.

Actually, I think everything Elon said appears to be technically true--the car wasn't fully charged, and the driver took a detour. What he didn't say was whether or not the car can be normally "partially charged" after a "charging complete" status posted on the car (yes? no?), whether or not the lost range due to a incomplete charge was material (the journalist claims the car reported sufficient mileage before the drive began) and whether the detour materially affected the result of the review (if it's really a two mile detour, the answer is probably "no"). Does all this add up to a "faked" review? Not sure, but I wouldn't blindly trust Elon's classification.

> What he didn't say was whether or not the car can be normally "partially charged" after a "charging complete" status posted on the car (yes? no?)

The reporter actually touched on this in his rebuttal. He claims that no one ever told him to set the charge mode to max range, instead of "normal," so the car stopped charging when it got to 90%. However, he stops short of actually saying that he wasn't aware of the existence of "max charge" mode.

He also says that nobody told him to have the car charge while he was sleeping. Personally, I see this as complaining that your interior is wet because Ford didn't tell you roll the windows up before driving through the car wash--it ought to be common sense. Maybe it can be forgiven in a new and unfamiliar car, but it sure seems like the author didn't go out of his way to figure out how to operate it.

In the HN thread about the original story, someone noted that he only mentioned he speed after Tesla told him to slow down. It seems that that was a rather prescient comment in light of the recent claims by Musk that the author was driving well in excess of the speed limit and driving briskly.

You ask if the review is "faked." I don't think "faked" is quite the right word, but there were certainly some omissions in the original write-up. By omitting certain facts, I think the author presented the story in a way that isn't the shining pinnacle of honesty. The Model S may very well not be capable of successfully making the trip the author wanted to make. However, if that is the case, the author shouldn't need hide behind technicalities and loaded language in his review.

Yes, he has the logs. So lets see them. This has now been going on for longer than the drive itself. How much time do they need to review the logs ?

This has me scratching my head.

Since the story blew up, the NYT has had time to get their graphics department to draw a nice infographic (that was up last night) and now a written defence by the original author of the story.

If Tesla's logs are such irrefutable evidence of fakery, why's it taking so long to post it online?

I'm sure a nice infographic and a simple article that requires zero research are significantly easier to whip up than a good analysis, presented in a way understandable to people, of a detailed set of logs.

What, so Musk couldn't just give out the raw logs for the nerds to digest with the message that a more consumable version was on its way? It takes less than 2 minutes to post the raw logs.

In any case, for Tesla to present an analysis without also giving the raw logs removes any ability for the public to fact check the analysis.

You're assuming that it's appropriate for Musk to release the raw logs. I have no idea what's actually collected, but it's quite possible that the logs contain data that either should not, or cannot, be released.

As a simplistic example, just because the journalist agreed to let Tesla collect the logs from his trip doesn't mean that the journalist agreed to let the precise details about his driving habits get released to the public. And there could be more than just privacy concerns. Maybe the logs contain data related to Tesla trade secrets.

If they can't release everything related to that trip, then it leaves Tesla wide open to the question that they just doctored up the analysis of the trip.

In the same way we shouldn't take NYT's story at its word, we should not take Tesla at their word.

If you're going to be that paranoid, what's to stop you from accusing Tesla of doctoring the raw logs themselves if they do get released?

Nothing will.

But look at the posts in the threads related to this story. There's plenty of paranoia going around, with the automatic presumption that the NYT had some axe to grind with the original story or financial incentives to embellish the story.

I just think that any claims that Tesla makes deserves the same level of scrutiny that people here have been giving the NYT article.

Personally I would not expect them to release 'raw' logs. That would probably be confusing to most people and like many propriety logs contain lots of info that any geek bystander could not properly interpret fully accurately.

However, given the visibility of this I would expect that they would make releasing appropriately annotated logs a priority.

I doubt that they'll release the logs too.

But when you are basically screaming out "You're lying, and I have proof of it" and then take your time delivering the proof, it doesn't look very good in the court of public opinion.

Until the logs are released to the public, it is a case of "he said, she said."

> And consider the consequences if Musk reported something different than the records show -- and Musk knows this. Why would he knowingly sabotage his own company?

Wouldn't a privileged DBA be able to make the data look like just about anything? Also, the location logging system as a whole isn't necessarily infallible.

And if the NYT ever got a hold of any reasonable evidence of doing so, at best (for Tesla) it'd ruin Musk's credibility and at worst they could turn it into libel charges of their own.

Plus being publicly traded, I believe the SEC might have words on the subject.

> Plus being publicly traded, I believe the SEC might have words on the subject.

The SEC would only get involved in the event of an IPO, an insider-trading scandal, or some other market-related event, not a dispute about accurate media reporting.

Indeed, if that location data's GPS, it can get pretty distorted in the city. Especially if there's a lot of thick cloud coverage.

The Tesla dash runs Android, doesn't it? I wonder if it can use wifi & cellular data to help correct GPS like most mobile devices can.

> The Tesla dash runs Android, doesn't it? I wonder if it can use wifi & cellular data to help correct GPS like most mobile devices can.

Unless the car is under a wet forest canopy, the GPS position will be more accurate than the other ways of determining position.

> Indeed, if that location data's GPS, it can get pretty distorted in the city. Especially if there's a lot of thick cloud coverage.

No, city GPS readings tend to be either plausible or nonexistent. Most receivers are able to detect absurd apparent positions created by building reflections (they do it by comparing different combinations of satellites and accepting the majority result).

Wet forest and country road, yes -- big problem for GPS. City driving, no, no really. And cloud cover doesn't significantly affect GPS performance:

http://gpsinformation.net/gpsclouds.htm

Reply on news.ycombinator.com