I had not read that Graham post before. Interesting ..
Keep in mind the term Public Relations was though to have been coined by Edward Berneys. He's said on camera that the US Government realized propaganda worked well during WW2, but propaganda had become a bad word. So he created the US Department of Public Relations. It's literally a weasel word for propaganda.
I should've been more descriptive. In general, it's hard to know who's "ahead" as neither of them has launched anything. However, Google has spent more capital over a longer time on it. This piece though seems mostly to be PR covered up as journalism (aka the submarine).
The headstart that both Tesla and Google have is test data. Tesla has more, because they've collected data from production cars from the get go. Google likely has more mature software (pure speculation), and it has comprehensive mapping data, but it doesn't have 100k customers providing them with useful real-world information. It does, however, have millions of Android users, but that's not the same. I believe all new Tesla vehicles send back autopilot data to the mothership.
There's some interesting comparisons. On the one hand you have Google, which is primarily a software company with hardware interests. Then you have GM which is an auto company with software interests. And you have Tesla which positioned itself in the middle, aiming to both build cars and software.
As a result, Google has a lot of core software expertise and infrastructure (particularly for machine learning). GM has the production capacity and the ability to roll out features to potentially millions of new vehicles. Google had to buy in automotive people. GM had to buy in software people. Tesla had to buy in both, but they did it from the very beginning with clear goals in mind.
From an outsiders' perspective, building a car that can drive by itself isn't terribly difficult nowadays. If you have the cash, you can buy a team that will build it. The challenge is getting from that stage to a car that can reliably drive better than a human in all driving conditions and deal with unexpected events. That will take much more time, and much more data.
There are other aspects like getting the sensors down to a price that people can actually afford.
I quit buying American cars because they are always a "me too", this seems like another case. Every American made car that I have boughten in the last 10 years has crapped out way before any Japanese car of similar value. The one caveat I have is Tesla, I would gladly buy one of those.
I think it will be interesting to see a head-to-head comparison between the Chevy Bolt and the Tesla Model 3. Even though the Model 3 will be more expensive they seem to be comparable, aesthetics aside.
I understand that quality control, including fit and finish, has been steadily improving.
An electric car such as the Model S has some intrinsic advantages over ICE cars in reliability in that the drive system is small and relatively uncomplicated. I think there is likely some bias in the perception of Tesla's reliability in that there is a great deal of attention on the company so problems, no matter how small, are more likely to be reported and commented upon than with more established car makers.
There are a few specific examples that I'm aware of. For example the Model S pop-out door handle was redesigned and has been steadily replaced, and the Model X doors had no end of problems until they apparently sorted them out.
So are German cars. Audi is a more expensive VW, Skoda and Seat are cheaper ones but the models are very much similar. The new Seat Leon looks like a VW golf now and so does the new Skoda Fabia. Mercedes is a more classy BMW. Opel was owned by GM before being bougt by Peugeot. Only Porsche, Audi and BMW make some unique sports cars.
I would rather buy a Japanese car any day. They're more reliable and you get more car features for the same amount of money.
the car market is definetly oversaturated
they seem to be sustaining this post-equilibrium state with government subsidies and 'strategic acquisition' i.e. monopolization. wich hopefully gets the profit margins in order, but at leasts aids in more governments-gibs (via stronger coordination in their efforts of leveraging power/lobbying, having more leverage)
also chinese are making investments (like buying daimler/mercedes-benz) that would be -ev in a vacuum. they are doing it trying to find a way into the market after beeing cockblocked EU & consorts. and for the tech I assume.
wouldnt buy a tesla though. that level of computry and no FOSS, or at least OSS, forms too high a risk as is. plus there is sending a signal that we want/require these things, so we can actually have them instead of getting cucked all the time.
also am I uninformed or do they loose like 90+% value when the battery dies (8y warranty)
btw data collection is fine within certain limits. the reality is many people do need privacy with all the persecution going on and rising (subverted alphabet soup agencies, militant semi-organized communists or good old moderate islamists beating up right-wing politicians)
so it is necessary to have that option. would you buy a car that requires you to censor yourself on certain topics lest you may be facing zelot vigilantes or worse? its not like the 'as long as you have nothing hide..' slogan is anything other than propaganda/dysinfo
I understand the realities of the market, competition etc. so I dont blame tesla for everything. still not a product I want.
That's such a useless article. I, too, want be the leader in driverless cars. And much like GM, I'm not even close to being the one who will do it! Though, in all fairness, GM is a lot further along than I am.
I don't understand the negativity for GM/Ford on HN. I'd say with cruise / lyft partnership / sidecar acquisition they're way further along than anyone here seems to give them credit for.
It seems that the sentiment for more alternative and "disruptive" ideas or businesses are the most welcomed here. It's a shame that people are missing indicators that the enterprise is < 10 years away from disrupting traditional software startups, software is becoming a commodity.
they are trying to buy their way into innovation
however they dont have any attractor for that of their own - not a real economic opportunity behind them. just a bunch of leftover money. 'talent acquisition' looks too much like yahoo. companies need an adhesive, typically an enormously strong forcing function caused by economic currents. it can be other things, like a magnetic leader etc.
GM has none of that, only gubermint bailout money
so while that is going to get somewhere (since they have a lot of it) I would assume they will still be outcompeted by other companies with stronger attractors (econ opportunities, PR, ..) and movers (like mission/aspriations, inspriation, arbiters of quality, ..)
and of course there is the negative sentiment about it beeing illegitimate, unearned, perhaps undeserved if epistemology swings that way, because of the above
People in the Bay Area (and many people on HN fit this demographic culturally, if not literally) tend to be ethnocentric. I have met so many people that won't take a company seriously unless it is headquartered locally, including at least 2 people who didn't believe me that Amazon was based in Seattle. There is a huge misunderstanding of the auto industry and since people do not understand it they tend to discredit it. This hubris is largely why Google and Apple attempted to build their own cars, and after failing decided to pursue other routes (Waymo partnering with Chrysler and Apple doing who knows what). There is an incredible arrogance in the Bay Area, and denizens tend to think they are smarter and able to disrupt the rest of the world (much like this https://xkcd.com/1831/).
As HN becomes more popular it inevitably becomes more cliquish. While I commend dang on his typically excellent moderation, it is inevitable that with more people and a voting system that a hivemind effect will develop.
Since I'm sure people will criticize me for being new here, this is not my original account and I have been here for a while.
Well they hired people and then fired them, so whatever you want to call that. I have a friend that was at Google and a friend that was at Apple, each was laid off because they no longer want to manufacture cars, which is why they were hired. Also Google actually did manufacture those weird small cars that they've been using for some time, and decided they don't want to mass manufacture them.
It's very, very difficult to build a successful car company, especially one that can produce affordable cars at scale.
The technology for self-driving cars will become commoditized, just as the technology behind smartphones did. When that happens, it will be the companies like GM, Ford, VW, and Toyota that produce self-driving cars that people can afford at large scale.
Tesla wants to join that club, but their production scale is not anywhere near that of GM or Ford, and they are about to learn that the biggest obstacle to building millions of cars in multiple factories is not technology and production processes -- it is dealing with worker unionization and the local/national politics of the places they try to build factories in. This is something the large players like GM and Ford are good at, after decades of experience.
Tesla is also getting an easy ride because it has a lot of very happy shareholders right now. Imagine trying to surmount those challenges while angry activist shareholders are trying to get you to split the stock, or forcing out your CEO when the stock price is flat.
I think Ford and GM are bargain stock deals right now for people who want to invest in self-driving cars. Ford and GM are may not be the first to develop that technology, but they are probably going to build more self-driving cars than anyone else once the technology becomes available.
I really think we're a lot further away from self driving car tech than people think. The fact that every company is building totally closed proprietary systems should also be of great concern. What if Waymo implements a critical safety feature and patents the software/algorithm? Are we going to see Here or GM touting that their algorithms are safer in their adverts? Are we going to have mini-DARPA challenges for each manufacturer, similar to crash safety testing?
I also just feels like it's a lot of wasted money. GM killed the Streetcar in America and we've been left with a country without any rail infrastructure. Self driving cars make sense in Europe where they could solve last mile problems, but in the US, it's not going to solve any of the major transportation issues: namely congestion and gridlock. There is no way a fleet of self-driving cars, no matter how advanced, can even hope to match the carrying capacity of a tram or train:
There is even a reasonable argument to be made that self driving cards will increase traffic and deadlock.
For example, imagine driving in to the city for some business meeting and then rather than worry about the hassle of finding a parking spot you just let the car drive about by itself for a couple of hours. Now multiply that by however many hundred (or thousand) people have the same idea.
Transportation is one of the few contexts where the US has adopted the socialist system of allocation by waiting list, rather than price.
Congestion used to be the one inconvenience rich people couldn't buy themselves out of. I can understand the impulse to keep it that way, though the benefits of congestion pricing probably do outweigh the symbolic problems.
>What if Waymo implements a critical safety feature and patents the software/algorithm?
Maybe I'm naive, but is this really different than today? Let's say Volvo implements some new kind of crumple-zone tech that drastically reduces fatalities in head on collisions, and patents the tech. Is this a big ethical problem?
I think the distinction here is that failure conditions in current vehicles are almost entirely human caused, so that new safety tech only matters after a human has already made a mistake. With autonomous cars, humans aren't making mistakes anymore, software is.
So essentially, we're looking at transitioning from protecting people from themselves, to protecting people from the machines. Which is something we FEEEL very differently about, even if the big picture isn't all that different. I mean, it's certainly something that changes the dynamics of liability, that just a matter of of who's paying the insurance bill.
I doubt people will be able to own self driving cars. They are going to need the maintenance schedules of small aircraft to be safe. It's more likely they'll only be available as a service or taxi.
I also think this is a big reason for the apprehension of self driving cars. While some advocate safety, people are more than familiar with software bugs in our phones, computers and web apps.
Sure there is a difference. The software in mission critical applications like pacemakers, aircraft navigation, etc. has to be thoroughly tested and designed to be either failure proof or failing in a safe way or having a redundant system. But will the products Uber and Google develop have the same level of testing Boeing puts into aircraft or SpaceX puts into rockets?
We have a historical record of this - a company invented the three point seat belt and felt it was too important to keep to themselves for the safety of car users around the world and freely licensed it so that every car company could incorporate it. That company was Volvo.
https://priceonomics.com/volvo-gave-away-the-most-important-...
Not hating on your comment - it's actually rather interesting and informative. What is also interesting to me is how you wrote it:"That company's name? Xxxx" after a description. This reads like click bait. I've noticed an uptick in people writing like this. I wonder if this a trend others are seeing?
That's one of the oldest rhetorical devices in the book. It's a common technique of good storytelling. Clickbait is annoying because of its context, but the reason it works is because those techniques stimulate human curiosity, and increase their enjoyment of the story.
That's true but the OP gave a mechanical example, too! :) On the other hand my out of loop observation is we're missing some tech revolution/discovery to make SDC possible because every single large car company and some big name tech companies are on the hunt and it feels like no one is close, though I have no concrete data on that, just watching the Waymo cars drive around my home.
If miles are cheap it will likely increase vehicle miles.
I guess in many areas public transport doesn't make much sense to begin with (low density).
In higher density areas, it can impact the decisions people make. A household having 1 car or no car makes it an easier decision to take public transport if it is cheaper than hailing a self driving taxi (or faster during a rush hour or whatever). It makes it easier to bike to work without worrying about the weather. And so on.
> I guess in many areas public transport doesn't make much sense to begin with (low density).
I really hate this argument. America use to have tons of rail. People today say "America is too spread out," but that didn't just happen. It grew that way.
Cities like Jena, Germany have ~100,000 people, yet it has 4 tram lines and three intercity stations. Chattanooga, home of the famous Choo-Choo song, has no Amtrak lines at all (where once it had street cars and was a major intercity rail stop).
In America if we started building rail, we'd watch suburbs collapse. Parking lots are huge blocks of lost real estate. Once you have rail and 1/3 of the lots are continually empty, you'll see stores selling that space to developers. Suburbs can being to collapse, simply because of the insane amount of parking lots we have.
I think Seattle will be good example of if this theory holds true as the ST3 expansion will push deep into a lot of suburbs.
49 comments
[ 6.4 ms ] story [ 107 ms ] threadThere's nothing in the article that others haven't already done.
http://www.paulgraham.com/submarine.html
Keep in mind the term Public Relations was though to have been coined by Edward Berneys. He's said on camera that the US Government realized propaganda worked well during WW2, but propaganda had become a bad word. So he created the US Department of Public Relations. It's literally a weasel word for propaganda.
There's some interesting comparisons. On the one hand you have Google, which is primarily a software company with hardware interests. Then you have GM which is an auto company with software interests. And you have Tesla which positioned itself in the middle, aiming to both build cars and software.
As a result, Google has a lot of core software expertise and infrastructure (particularly for machine learning). GM has the production capacity and the ability to roll out features to potentially millions of new vehicles. Google had to buy in automotive people. GM had to buy in software people. Tesla had to buy in both, but they did it from the very beginning with clear goals in mind.
From an outsiders' perspective, building a car that can drive by itself isn't terribly difficult nowadays. If you have the cash, you can buy a team that will build it. The challenge is getting from that stage to a car that can reliably drive better than a human in all driving conditions and deal with unexpected events. That will take much more time, and much more data.
There are other aspects like getting the sensors down to a price that people can actually afford.
https://electrek.co/2017/04/06/tesla-model-3-average-sale-pr...
http://www.consumerreports.org/cars-tesla-reliability-doesnt...
https://www.fool.com/investing/2017/03/07/repairing-my-tesla...
I understand that quality control, including fit and finish, has been steadily improving.
An electric car such as the Model S has some intrinsic advantages over ICE cars in reliability in that the drive system is small and relatively uncomplicated. I think there is likely some bias in the perception of Tesla's reliability in that there is a great deal of attention on the company so problems, no matter how small, are more likely to be reported and commented upon than with more established car makers.
There are a few specific examples that I'm aware of. For example the Model S pop-out door handle was redesigned and has been steadily replaced, and the Model X doors had no end of problems until they apparently sorted them out.
I would rather buy a Japanese car any day. They're more reliable and you get more car features for the same amount of money.
wouldnt buy a tesla though. that level of computry and no FOSS, or at least OSS, forms too high a risk as is. plus there is sending a signal that we want/require these things, so we can actually have them instead of getting cucked all the time. also am I uninformed or do they loose like 90+% value when the battery dies (8y warranty) btw data collection is fine within certain limits. the reality is many people do need privacy with all the persecution going on and rising (subverted alphabet soup agencies, militant semi-organized communists or good old moderate islamists beating up right-wing politicians) so it is necessary to have that option. would you buy a car that requires you to censor yourself on certain topics lest you may be facing zelot vigilantes or worse? its not like the 'as long as you have nothing hide..' slogan is anything other than propaganda/dysinfo I understand the realities of the market, competition etc. so I dont blame tesla for everything. still not a product I want.
and of course there is the negative sentiment about it beeing illegitimate, unearned, perhaps undeserved if epistemology swings that way, because of the above
As HN becomes more popular it inevitably becomes more cliquish. While I commend dang on his typically excellent moderation, it is inevitable that with more people and a voting system that a hivemind effect will develop.
Since I'm sure people will criticize me for being new here, this is not my original account and I have been here for a while.
not sure when you think either attempted this...
http://www.popularmechanics.com/cars/technology/a24274/googl...
It's very, very difficult to build a successful car company, especially one that can produce affordable cars at scale.
The technology for self-driving cars will become commoditized, just as the technology behind smartphones did. When that happens, it will be the companies like GM, Ford, VW, and Toyota that produce self-driving cars that people can afford at large scale.
Tesla wants to join that club, but their production scale is not anywhere near that of GM or Ford, and they are about to learn that the biggest obstacle to building millions of cars in multiple factories is not technology and production processes -- it is dealing with worker unionization and the local/national politics of the places they try to build factories in. This is something the large players like GM and Ford are good at, after decades of experience.
Tesla is also getting an easy ride because it has a lot of very happy shareholders right now. Imagine trying to surmount those challenges while angry activist shareholders are trying to get you to split the stock, or forcing out your CEO when the stock price is flat.
I think Ford and GM are bargain stock deals right now for people who want to invest in self-driving cars. Ford and GM are may not be the first to develop that technology, but they are probably going to build more self-driving cars than anyone else once the technology becomes available.
"Wants to" doesn't win pennants.
I also just feels like it's a lot of wasted money. GM killed the Streetcar in America and we've been left with a country without any rail infrastructure. Self driving cars make sense in Europe where they could solve last mile problems, but in the US, it's not going to solve any of the major transportation issues: namely congestion and gridlock. There is no way a fleet of self-driving cars, no matter how advanced, can even hope to match the carrying capacity of a tram or train:
http://penguindreams.org/blog/self-driving-cars-will-not-sol...
For example, imagine driving in to the city for some business meeting and then rather than worry about the hassle of finding a parking spot you just let the car drive about by itself for a couple of hours. Now multiply that by however many hundred (or thousand) people have the same idea.
Congestion used to be the one inconvenience rich people couldn't buy themselves out of. I can understand the impulse to keep it that way, though the benefits of congestion pricing probably do outweigh the symbolic problems.
Maybe I'm naive, but is this really different than today? Let's say Volvo implements some new kind of crumple-zone tech that drastically reduces fatalities in head on collisions, and patents the tech. Is this a big ethical problem?
I think the distinction here is that failure conditions in current vehicles are almost entirely human caused, so that new safety tech only matters after a human has already made a mistake. With autonomous cars, humans aren't making mistakes anymore, software is.
So essentially, we're looking at transitioning from protecting people from themselves, to protecting people from the machines. Which is something we FEEEL very differently about, even if the big picture isn't all that different. I mean, it's certainly something that changes the dynamics of liability, that just a matter of of who's paying the insurance bill.
I also think this is a big reason for the apprehension of self driving cars. While some advocate safety, people are more than familiar with software bugs in our phones, computers and web apps.
Sure there is a difference. The software in mission critical applications like pacemakers, aircraft navigation, etc. has to be thoroughly tested and designed to be either failure proof or failing in a safe way or having a redundant system. But will the products Uber and Google develop have the same level of testing Boeing puts into aircraft or SpaceX puts into rockets?
I guess in many areas public transport doesn't make much sense to begin with (low density).
In higher density areas, it can impact the decisions people make. A household having 1 car or no car makes it an easier decision to take public transport if it is cheaper than hailing a self driving taxi (or faster during a rush hour or whatever). It makes it easier to bike to work without worrying about the weather. And so on.
I really hate this argument. America use to have tons of rail. People today say "America is too spread out," but that didn't just happen. It grew that way.
Cities like Jena, Germany have ~100,000 people, yet it has 4 tram lines and three intercity stations. Chattanooga, home of the famous Choo-Choo song, has no Amtrak lines at all (where once it had street cars and was a major intercity rail stop).
In America if we started building rail, we'd watch suburbs collapse. Parking lots are huge blocks of lost real estate. Once you have rail and 1/3 of the lots are continually empty, you'll see stores selling that space to developers. Suburbs can being to collapse, simply because of the insane amount of parking lots we have.
I think Seattle will be good example of if this theory holds true as the ST3 expansion will push deep into a lot of suburbs.