Live data from Hacker News

Master Plan, Part Deux

tesla.com

651–660 of 705 posts

Re: Master Plan, Part Deux

#651

Earlier quoted context omitted.

That's his sales market. Tesla wants fanboy buyers who will be upgrading within five years and that's not a bad thing. They are such a small player that they can do this. I used to lease SUVs on 2-year lease. They had massive incentives and very high residuals because the manufacturers realize they can sell you 5 cars every decade if you are loyal. I did it because the savings were awesome, but I eventually stopped b…

I'm not sure I understand. Are you saying it cost less to lease 5 cars over a 10-year period than to buy one and hold it?

Not at all, but it was incredibly affordable considering that in 10 years I never had a car older than 2 years. There is zero maintenance other than a few oil changes. Like I said, the residuals were usually so high that the buy-out was not realistic. That kept the lease rate very low and made it unthinkable to keep it at the end.

Re: Master Plan, Part Deux

#652
post #606

Earlier quoted context omitted.

AbstractTeslaIndustry is my personal convention in this case.

Aha! But how would you create a TeslaFactoryFactoryFactory?

AbstractTeslaEconomy. Then it'd be eventually AbstractTeslaNation or AbstractTeslaCivilization depending upon if the class is a singleton.

Also, I would probably be evaluating a DI framework long before thinking of anything beyond an Industry. Even an EJB sounds great by that point.

Re: Master Plan, Part Deux

#653
post #639
post #560

Earlier quoted context omitted.

I live 8 miles from work. "Limiting my range" from 200 miles to 150 miles between daily charge ups is irrelevant.

Not to mention that battery tech will hopefully improve once everybody relies on battery powered cars, and some currently infeasible tech becomes cost-effective due to large scale deployments.

The rate of improvement in battery tech has been painfully slow.

Re: Master Plan, Part Deux

#654
post #501

Earlier quoted context omitted.

What does that data give you? Raw sensor data + steering angle + acceleration/braking is not particularly useful in of itself. Nobody is trying to shove large amounts of data through a blackbox model and directly trying to predict car controls from images. This is not MobileEye's approach and almost certainly isn't Tesla's approach. The bottleneck is data that is decomposed and labeled in greater detail (lane marking…

I recommend watching this [0] ted talk on the subject. In it, you'll see that Google, last year, already had solved bicyclist hand signs, lane markings, unforeseen obstacles, etc. And they did it __without__ human annotators. Just some straight up convolution neural networks. [0]: https://www.ted.com/talks/chris_urmson_how_a_driverless_car_...

You're mistaken. Convolutional neural networks require annotated data. It's a TED talk; they're obviously not going to mention the very boring details of their annotation protocols.

Re: Master Plan, Part Deux

#655

Earlier quoted context omitted.

Why would it make any sense to account for the effort needed to produce oil from high-entropy substrates and an energy source? If you're going that far, why not take it to its logical conclusion and ask how much it would cost to create a barrel of crude oil starting from nothing at all, creating your own Big Bang and going from there? As Sagan said, if you wish to make an apple pie from scratch, you must first invent…

Looking at the actual cost of formation of fossil fuels makes sense from a total accounting standpoint. That's a depletion of an existing capital stock -- a bank account, if you will. Might want to talk to the good citizens of Nauru about their experience. And actually, tracing back resources to their antecedents is useful -- there are compounds which formed through biological activity (fossil fuels, limestone, most…

That's all very interesting, and I totally agree that when you're extracting a resource faster than it's being created, you need to somehow account for the fact that the available quantity is finite.

But I don't see why looking at the energy used to create the resource, or the cost of creating it from scratch, is useful. All that matters today is what's there and how much we use. For example, consider oil versus uranium. In terms of usable energy content, there's a lot more uranium than oil on the planet. I haven't run the numbers, but I'm guessing that stellar nucleosynthesis of heavy elements like uranium is way less efficient than turning CHON into oil with photosynthesis and time. Certainly, if we were creating this stuff today, it would be way cheaper to create a megajoule of oil than a megajoule of U235. But, who cares? We have more uranium, so it ought to be cheaper per megajoule, all else being equal (which it very much isn't, of course).

You say that we're consuming about five million years of fossil fuel production per year, and that there's about 200-300 million years total. The rate is basically ignorable compared to usage, so what really matters is reserves divided by consumption rate. The relevant number is 40-50 years of reserves at current rates. For example, say it turns out that the geologists and petrochemists got it wrong, oil is actually much more energy intensive to produce naturally, and current reserves actually represent 400-600 million years of accumulation. The energetic cost accounting changes dramatically; each barrel of oil now represents twice as much input energy! But our economic situation today doesn't change one bit.

Taking your inheritance analogy, it doesn't matter how much money your rich family line started with, all that matters is how much is there now. It doesn't make a bit of difference to your spending whether your $10 million inheritance is the paltry remains of a $10 billion fortune, or the amazing end result of a $10 investment.

The whole business of input energy and cost to recreate just seems so very arbitrary. For example, why are you counting the cost of the oil, but not the cost of the oxygen you use to burn it? Both parts are necessary, and the fact that oxygen is ubiquitous and easy to access just makes it even more of a subsidy. Shouldn't we be counting not only all the fossil energy needed to create the oil we burn, but all the fossil energy needed to crack all that oxygen out of CO2?

Re: Master Plan, Part Deux

#656

Earlier quoted context omitted.

Low density remote developments or just typical suburbs cause higher pollution, higher service costs, segregation, hurt smaller businesses and have a negative impact on health. Modern urbanism is very opposed to urban sprawl because of those and many other reasons. Transport is just one aspect of it.

I'm curious about your "have a negative impact on health" approach. The density of a typical city is orders of magnitude greater than anything that our ancestors have seen in nature; and, historically, cities have been a breeding ground for all kinds of diseases (because sanitation is harder and spreading is easier). Can you clarify what exactly you meant? I'm not sure that higher pollution is innate in lower densiti…

Walking is good for your health. It's basically impossible to walk anywhere in the suburbs. Many don't even have sidewalks.

In a car oriented suburb where you have to drive to literally every location you have to try extra hard to get even a basic level of exercise.

Re: Master Plan, Part Deux

#657

Earlier quoted context omitted.

>Without need for parking, commercial districts could get denser and denser, and people could have nice large homes far away for cheaper than ever. If people are living farther away from their jobs, you're likely to get more congestion, not less.

If this scenario has a majority of cars autonomously driven, congestion is less of a problem. The cars would cooperate to solve congestion issues that individual human drivers struggle with. Also, as a passenger you're probably sleeping or being productive within the car, so the increased commute time is less of a problem.

> The cars would cooperate to solve congestion issues that individual human drivers struggle with.

Does this effect overcome the added amount of cars on the road from people who can now drive that wouldn't be able to before? I'm skeptical.

Re: Master Plan, Part Deux

#658

"I should add a note here to explain why Tesla is deploying partial autonomy now, rather than waiting until some point in the future. The most important reason is that, when used correctly, it is already significantly safer than a person driving by themselves and it would therefore be morally reprehensible to delay release simply for fear of bad press or some mercantile calculation of legal liability." That's a hell…

Here's a probably equally badly made statistic. Tesla estimates their cars had driven 130 million autopilot miles before the fatality. Some googling suggests the first non-autopilot Tesla S fatality was in july 2014 and this graph ( http://insideevs.com/wp-content/uploads/2015/06/Tesla-Model-... ) suggests that was around 400 million miles driven. So the autopilot is 3.33 times 'worse'? Who knows, but if Tesla feels…

In other words, if you're a Tesla driver, you're safer driving the car yourself than driving it with autopilot.

Re: Master Plan, Part Deux

#659

Earlier quoted context omitted.

Cities already tend to have limited parking. We might see more parking structures located in cheap out-of-the-way locations rather than near places people want to go. And street parking could go away in favor of consolidated parking structures, freeing up road space for both car traffic (with extra lanes) and foot traffic (widen sidewalks). But I think it'll have an even larger effect on suburban areas. Shopping cent…

Let me spell out the argument more: Instead of 100 people owning 100 cars they drive 5% of the time, they could own 10 summonable cars that drive them around 50% of the time. The first scenario requires 95 parking spots. The second only 5. There is still as much traffic on the roads, and as much transport work being done, but most of the space now needed for parking can be put to other use.

Why do you need autonomous vehicle technology to realize this?

One could do this already by just taking taxis everywhere.

This obviously doesn't happen that much.

What is autonomous car technology going to change that will make this concept a reality?

Re: Master Plan, Part Deux

#660

Earlier quoted context omitted.

If this scenario has a majority of cars autonomously driven, congestion is less of a problem. The cars would cooperate to solve congestion issues that individual human drivers struggle with. Also, as a passenger you're probably sleeping or being productive within the car, so the increased commute time is less of a problem.

> The cars would cooperate to solve congestion issues that individual human drivers struggle with. Does this effect overcome the added amount of cars on the road from people who can now drive that wouldn't be able to before? I'm skeptical.

That was my point. If I buy a van that drives itself I can tolerate much longer commutes - I'll just put a bed in the back and sleep my way to work. So what if it takes three hours.

Incidentally, I was talking with a government transportation engineer recently, and that's what keeps them up at night - trips that are too long or just too inconvenient today become doable with self driving cars.

Post reply on HN