Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Tesla Q1 2021 Results

tesla-cdn.thron.com

Post reply on HN
Print view View original

171–180 of 591 posts

Re: Tesla Q1 2021 Results

#171

Post by nimithryn »

feifan wrote: post #89
anonu wrote: post #50

> We believe that a vision-only system is ultimately all that is needed for full autonomy. Our AI-based software architecture has been increasingly reliant on cameras, to the point where radar is becoming unnecessary earlier than expected. As a result, our FSD team is fully focused on evolving to a vision-based autonomous system and we are nearly ready to switch the US market to Tesla Vision. This is a huge gamble IM…

I'm not entirely convinced it's a good idea, but one convincing argument I've seen is that sensor fusion is hard and it's not a priori clear which sensor to believe when they provide conflicting information. It's entirely possible that they get better predictive accuracy from vision-only data then in the analogous situations with vision+radar data.

I don't find this argument convincing at all. Just ignore the lidar in situations where lidar and vision conflict and you have the original decision.

They probably started without lidar due to the cost (Tesla is one of few companies actually selling cars to customers, who are sensitive to price). The price of lidar has come down, but I would guess it requires a fair amount of expensive rearchitecture to add it in now, and they haven't bitten the bullet yet.

Re: Tesla Q1 2021 Results

#172

Post by clouddrover »

dangus wrote: post #10
systemvoltage wrote: post #4

Earlier quoted context omitted.

What’s amazing is that just 5 years ago, it seemed like a distant goal. People were skeptical of Tesla in 2015 (rightfully so). Almost went bankrupt in 2017. They dug themselves out of there despite of Elon’s PR shenanigans. Pretty amazing I think! Now about that stock price, I’m not so sure. It’s like next 20 years are baked into $735/share price now.

The inflated stock price is easy to explain: intangibles are part of the balance sheet. In my guesstimation, Tesla will be making electric cars with a range to dollar ratio that can’t possibly be matched by incumbents for the foreseeable future. VW and the rest are more capable of producing a refined car, but (again, my opinion) Tesla’s EV advantage is similar to the big three’s various inertial advantages that they…

> Tesla will be making electric cars with a range to dollar ratio that can’t possibly be matched by incumbents for the foreseeable future

That isn't true today, never mind the future. If you want range to dollars then the Chevy Bolt is hard to beat. Especially with deals like this:

https://www.cnet.com/roadshow/news/chevy-bolt-ev-available-f...

It's also important to consider real world miles versus paper miles. Tesla's EPA range figures are inflated. On paper, the Tesla Model Y should have more range than the Ford Mustang Mach E. But in real world driving the Mach E has the greater range:

https://www.youtube.com/watch?v=lSmSiOo-v8s

Re: Tesla Q1 2021 Results

#173

Post by m463 »

nick_kline wrote: post #150
m463 wrote: post #67

Earlier quoted context omitted.

Now that you mention it - they're almost replicating the dealer experience. In the beginning Tesla's selling point was that they were not a dealer, you go to the showroom and there is no pressure because everyone pays the same price. Except with their pricing monkey business, you're replicating the same high-pressure "buy now before the price goes up"... for everyone. Aren't they over the hump? Can they get away from…

Every 2 or 3 months it changes for everyone when they +/- it by $500. It's not a negotiation. It is surprising to me they do this, I cannot understand the tweaking.

I was looking at a used car, and it changed price several times a week. You might also look at the experiences of used tesla buyers. The price isn't the only thing they've played with, they've been doing the same sorts of things with features and even warranty coverage, all of which have in general gotten worse.

Re: Tesla Q1 2021 Results

#174

Post by judge2020 »

systemvoltage wrote: post #4
11thEarlOfMar wrote: post #2

"Model 3 was the best-selling premium sedan in the world, outselling long-time industry leaders such as the 3 Series and E Class. This demonstrates that an electric vehicle can be a category leader and outsell its gas-powered counterparts." I find this to be a major milestone in the transition to BEV. This is an achievement.

What’s amazing is that just 5 years ago, it seemed like a distant goal. People were skeptical of Tesla in 2015 (rightfully so). Almost went bankrupt in 2017. They dug themselves out of there despite of Elon’s PR shenanigans. Pretty amazing I think! Now about that stock price, I’m not so sure. It’s like next 20 years are baked into $735/share price now.

Almost went bankrupt during the first Roadster deliveries in 2008 as well: https://www.bbc.com/news/business-43365710

Re: Tesla Q1 2021 Results

#175

Post by judge2020 »

azinman2 wrote: post #64
dangus wrote: post #10

Earlier quoted context omitted.

The inflated stock price is easy to explain: intangibles are part of the balance sheet. In my guesstimation, Tesla will be making electric cars with a range to dollar ratio that can’t possibly be matched by incumbents for the foreseeable future. VW and the rest are more capable of producing a refined car, but (again, my opinion) Tesla’s EV advantage is similar to the big three’s various inertial advantages that they…

So I just bought an ID.4 this weekend after seeing on HN on Friday that it launched in the US. I also test drove a Tesla Y just to double check Saturday morning before doing so. Tesla Y is for sure faster and the better drive. It's also like 10-15k more expensive and that's not even before VW's rebates. The interior was weird -- tiny headrests, tiny mirrors, and really poor back visibility. A giant screen in middle a…

> And that's driving the car? Scary!

The visualization happening there is more limited than what the car actually sees via the cameras, probably for performance reasons (it runs on a Intel Atom processor in the 3/Y) - you can see a more detailed representation on page 8 of https://tesla-cdn.thron.com/static/4E7BR9_TSLA_Q3_2020_Updat... and you can see more from public personalities with access to the beta https://www.youtube.com/results?search_query=tesla+fsd+beta

Re: Tesla Q1 2021 Results

#176

Post by Kranar »

t0mas88 wrote: post #29
11thEarlOfMar wrote: post #2

"Model 3 was the best-selling premium sedan in the world, outselling long-time industry leaders such as the 3 Series and E Class. This demonstrates that an electric vehicle can be a category leader and outsell its gas-powered counterparts." I find this to be a major milestone in the transition to BEV. This is an achievement.

The model 3 really isn't a premium sedan. It's more like a mid range Skoda than an Audi, with a cheap plastic interior and pretty bad build quality. They sell very well because it's cheap and in many countries you get a great tax deal on them making it even cheaper. Which I'm really happy for because EVs help against climate change. But I'm also very happy I cancelled my pre-order and don't have to drive a model 3.

What you're saying may be an important detail for car enthusiasts but that's mostly irrelevant given the context. From the business side of things what matters is that the Model 3 outsells all of its competitors at its price point.

If anything, the fact that a car that uses cheap plastic is outselling the car that doesn't use the cheap plastic at the same price is phenomenal.

Re: Tesla Q1 2021 Results

#177

Post by grecy »

nimithryn wrote: post #171
feifan wrote: post #89

Earlier quoted context omitted.

I'm not entirely convinced it's a good idea, but one convincing argument I've seen is that sensor fusion is hard and it's not a priori clear which sensor to believe when they provide conflicting information. It's entirely possible that they get better predictive accuracy from vision-only data then in the analogous situations with vision+radar data.

I don't find this argument convincing at all. Just ignore the lidar in situations where lidar and vision conflict and you have the original decision. They probably started without lidar due to the cost (Tesla is one of few companies actually selling cars to customers, who are sensitive to price). The price of lidar has come down, but I would guess it requires a fair amount of expensive rearchitecture to add it in now…

Elon has said many times they would use Lidar if they thought it was needed / beneficial. In fact he and the SpaceX team designed a Lidar into the Dragon spacecraft to help it dock with the ISS (which is obviously a HIGHLY sensitive operation with extremely strict rules), and he said it was perfect for that scenario.

He was asked point blank if they would use Lidar as Tesla even if they were free... and he said no. In their opinion it's simply not the right approach to generate photons and read them as they come back. It's preferable to simply read the photons that are already coming in.

Re: Tesla Q1 2021 Results

#178

Post by fy20 »

dawnerd wrote: post #119
t0mas88 wrote: post #68

Earlier quoted context omitted.

What you can buy instead is 99% based on personal preference, I cancelled the model 3 and got a BMW because I liked the few I had before. But my co-founder would never drive a BMW. He returned his second model S and got an Audi EV instead. Which is not a car I would pick. My point is that a model 3 isn't really a luxury car, it's an entry model and has the build quality and options that go with that.

So I don't get this. You bash the Model 3 for build quality but you support BMW which is notorious for having horrible reliability. I'd pick the car with just okay paint and maybe some not-so-perfect gaps over a car that I know is going to cost me thousands in maintenance.

OP could be in Europe. German premium cars are a lot cheaper to maintain there (still more than a Toyota though), and somehow seem to have less issues than the US.

Re: Tesla Q1 2021 Results

#179

Post by PhantomGremlin »

judge2020 wrote: post #170
TomVDB wrote: post #81

Earlier quoted context omitted.

Let's not forget impossibly advanced German car user interfaces, such as trackpad to gesture individual characters to spell a word!

For reference: https://youtu.be/_aV91djgwEc I'm astonished this made it to production. Red tape and bureaucracy means nobody is willing to push back against a feature that is bad and potentially dangerous (maybe not any more dangerous than touchscreens though).

In my experience working for a US subsidiary of a Japanese company, there's very little "push back" by underlings. That's simply not part of the culture.

As for peers, rather than underlings, the key word is consensus. Which is only achieved after endless months of interminable meetings during the day and interminable socializing during the evening. Maybe they finally reach consensus because they're worn out by all the discussion?

I hate touchscreens on the dash of moving cars. This interface is (perhaps?) worse, since the guy was making mistakes even in a stopped vehicle. But it would be interesting to try to use the interface while moving. One's hand is probably more stable down low, resting on the console, rather than extended out in space reaching for a touchscreen?

Either way, the interface is much worse than physical buttons on a console or dash.

Re: Tesla Q1 2021 Results

#180

Post by jrallison »

nimithryn wrote: post #171
feifan wrote: post #89

Earlier quoted context omitted.

I'm not entirely convinced it's a good idea, but one convincing argument I've seen is that sensor fusion is hard and it's not a priori clear which sensor to believe when they provide conflicting information. It's entirely possible that they get better predictive accuracy from vision-only data then in the analogous situations with vision+radar data.

I don't find this argument convincing at all. Just ignore the lidar in situations where lidar and vision conflict and you have the original decision. They probably started without lidar due to the cost (Tesla is one of few companies actually selling cars to customers, who are sensitive to price). The price of lidar has come down, but I would guess it requires a fair amount of expensive rearchitecture to add it in now…

"Just ignore the lidar in situations where lidar and vision conflict and you have the original decision."

Then why have lidar? If it's only purpose is to agree with vision or you just discard it, then why even have it?

Post reply on HN