Hacker News Reader

> legal reasons

Keep in mind that AARP is excited about driverless cars because for millions of seniors, losing their ability to drive means losing their independence. Also, by age 50, everyone knows somebody killed by a human-driven car. AARP's members vote at very high rates even in non-presidential elections, so they are a very powerful lobby.

http://www.aarp.org/home-family/personal-technology/info-201...

Not clear to me how a driverless car has any advantage for a senior vs. Uber or Lyft with a driver. In fact it seems less desirable, as the human driver could help the elderly passenger in and out of the vehicle, help with walkers or wheelchairs, etc.

The hope is that autonomous cars will be much cheaper once you take out the human element; when a taxi costs $30 for a routine trip, and Uber/Lyft still have to pay their drivers >minimum-wage, you cannot cater to the elder niche of doing groceries and making regular doctor trips, not to mention the useful aspect of 24/7 coverage.

There's also the psychological issues: elderly people don't want to be a 'burden' or rely on strangers, and autonomous cars neatly solve both issues.

It feels strange to see so many traditional backers for this tech. I've been hesitant to let myself get too excited for the possibility that driverless cars will reach the market when the technology is ready - instead of when it is deemed ready by regulators, lobbyists, and the gatekeepers who are backed up by legacy protectionist policies.

There seems to be so much positive traction among political and industry players that this might turn out to be one of those rare technologies who's progress doesn't get crippled by vested incumbents. But experience has taught me to be cautiously optimistic at best.

It does make sense though. The majority of people have driven a car and can relate to the hassles. Especially with drunk driving. Eliminating drunk driving is in everyone's best interest and hard to argue. Plus who would fight against self-driving cars? The only self-interest groups here would be the drivers themselves and even then I don't think it'll be a big enough lobby group.

Automobile manufacturers will see their market shrink considerably due to higher utilization from autotaxis.

Maybe a subscription model, now you can't pay once and maintain your car with your local mechanic for 20 years, it has to be done by the manufacturer and contains expensive software updates?

Well spotted.

It'll be "Cash for clunkers" redux. Much amplified is my guess. Human-driven cars will be ilegal on public roads in very short order.

There's one area where Lyft is ahead of Uber, the whole "we are nicer" thing discounted

The way cultural currents are going, there's a lot of room for companies to make marketing headway through "we are nicer" and "we are not douchebags" public images. (Hopefully, fairly accurate public images.)

I mean, when it's been years since Apple has been parodied with a "We're mad with power" tagline, and when public shaming has become a way of life online, it's about time for nice.

>the marketplace for on-demand drivers will become obsolete once autonomous driving becomes good enough to be adopted everywhere

I'm not holding my breath, it'll be a looong time (decades, plural) before driver-less cars work everywhere, in poor weather conditions, in areas with lots of cyclists and pedestrians, in developing nations and parts of developed nations with bad roads, bad or unenforced traffic laws etc.. Human drivers will have a reduced role, but they won't be "obsolete" any time soon.

Driver-less cars likely means people taking more trips, causing more congestion and more pollution on the roads where they can actually operate. I wouldn't be surprised if some cities ban them.

I think by then, electric cars will be cheaper to purchase and operate than gas vehicles, so the pollution won't be much of an issue.

But you are 100% right that congestion will be the problem. Cars simply do not scale well, as the amount of sqft/passenger/minute is huge. We seem to be unwilling to devote the huge amounts of land to highways that we would need to in order to make congestion go away. And in cities, intersections can only let so many vehicles through per minute.

As another commenter said, traffic is a good problem, however it's only good until it stops trips. It's great for people to be moving about, but when traffic prevents that, it's bad.

Autonomous drivers give people back their time in the car to do with as they please, but they will definitely increase congestion.

>the pollution won't be much of an issue

Depends. In St. Louis, for example, a fully electric vehicle produces the same pollution (via the power plant) as a 36mpg car [0]. That's about in line with a highly-rated compact like the Golf or Mazda3 these days.

Electric vehicles give us the capability to reduce emissions by building cleaner electrical infrastructure, but that does not mean we will do so.

[0] https://en.wikipedia.org/wiki/The_long_tailpipe#Carbon_footp...

In a few decades St Louis won't have such crappy energy sources, and if they do, their electricity prices will be sky high because they made poor decisions.

Additionally, pollution concentrations matter for cities, specifically dumping the exhaust out right next to people vs. dumping it at remote locations where it has more area to dilute before encountering people.

The long tail pipe hypothesis seems to be a bunch of people trying to desperately prove that electric cars are no better, but they have to resort to extreme cases to try to justify their positions. 36mpg which pollutants? Certainly not all them. CO2 from pure coal is something like a 50 mpg car. And as we get rid of these inferior and more expensive sources of electricity, electric cars will only increase their lead over gas vehicles.

A lot of congestion comes from cars looking for parking and cars not being shared with other people. The self driving cars are not going to be looking for parking, and can automatically scale what is on the road based on demand. They will also be more efficient with sharing a vehicle with multiple people, or have more specialized mini cars to scale trips properly.

Those self-driving cars are also going to be driving all over to pick people up, almost certainly more than is used for parking.

And when there's a surplus of self-driving cars, where do they go?

There will be an increase in miles driven per passenger mile in order to get the self driving cars to where they need to be, when they need to be.

Most people don't like to share rides with strangers. There may be a preference for smaller but more vehicles.

Smaller vehicles will be far more practical in an automated taxi model as well.

No one wants to own a single person car, because even a single person wants to be able to carry passengers if they need to. But lots of people will be making single person journeys all the time, so it makes sense for a taxi company to own them.

The added trips will offset the gains

Self driving mini busses semi-scheduled for commuters

Taking more trips to get to places they want to be. Traffic is a good thing, not a bad thing.

>Traffic is a good thing, not a bad thing.

Until your kid gets asthma so people can have the privilege of spending 2 hours in a car commuting.

Not to mention making the roads less usable for everyone who isn't in a car.

No, people taking trips is a good thing. Traffic is a mostly negative side effect of that, if they choose to travel by car.

> For technical and legal reasons, self-driving, on-demand vehicle services will be at least a decade out.

At least the legal reasons are now moot at least in Germany, as all the laws required for self-driving cars have already been passed.

All the harbors and airports have been using fully automated vehicles for transport on their area for quite a while now, and the recent showcase where autonomous trucks from all european car manufacturers drove to the netherlands has shown reliability on streets as well.

At least in the transport market autonomous cars are less than half a decade out.

There's a vast difference between cars that operate in a closed environment and cars that operate 'in the wild'.

Those areas are not more closed areas than the wild is. Everything can constantly change, people and cars can always appear anywhere in the path, etc.

So manufacturers already know under what conditions/standard they can put an unassisted self-driving car on the public roads, and it's the same as (or analogous to) the rule for automated port tools? What is that standard?

I don’t know myself, I’d have to look in the laws, but the rules were written because Mercedes and VW have now a decade of unassisted self-driving experience and plan to release the first fully self-driving vehicles for transport (trucks, etc) within of the next year.

"However, they still do not get the kind of data Google can get from their Maps and Waze (It's interesting to note that Lyft and Waze partnered) to enable their AI researchers."

Google is invested in uber, why would they not supply them the data they needed?

Uber is afraid of Google. That's why the maps in Uber's driver app can't plan a route according to traffic, they don't use google maps.

Uber is already integrated into Google Maps from the consumer side.

From the VC side, Google Ventures invested ~$250M in Uber a few years ago.

How is it they're afraid of Google?

Because Google routinely undermines their partners through internal productizing (see e.g. Flintstone/OnHub), and Google has a large self-driving car program. There's a clear path from there to a Google(/Alphabet)-owned ride provider regardless of investment in Uber.

Uber is building out their own map system so that they are no longer dependent upon a third party. It's already in production on a limited basis.

Isn't Google Ventures pretty isolated from Google?

Uber bought Bing Maps for $1.

This just isn't true in any way. Uber once upon a time was a bidder on Here Maps, which provided most of Bing's mapping data, but they lost to a consortium of German car makers (BMW, Daimler, Audi): http://www.wsj.com/articles/bmw-daimler-audi-agree-to-buy-no...

They acquired part of Bing's Mapping Assets alongside ~100 employees - http://techcrunch.com/2015/06/29/uber-acquires-part-of-bings... and http://www.wsj.com/articles/SB110643412133885342696045810783...

you mean $1B surely?

No. I can't name a source though. Basically Microsoft was hemmoraghing money on this unit and Uber took a chunk of staff, the tech, AND the operating expenses.

$1 doesn't make any sense. Perhaps you're leaving out some form of equity swap. Bankrupt companies get purchased all the time for millions of dollars and I can't imagine Microsoft's business team would allow such a shitty deal.

Bankrupt companies also routinely get purchased for $1, with the old owners just glad to get rid of outstanding liabilities.

It only makes sense if the unit was hemorrhaging money and had near iron clad contracts/termination penalties.

Google has the map data. Google are working on autonomous vehicles. When they get the latter, what is the need for Uber? The $250M would be an irrelevant sunk cost at that point.

I have no doubt about uber being around in 20 years, they're just too well established and understand where the industry is heading to become irrelevant.

Though what I do think will happen is an erosion of their lead where they go from the sole 800lb gorrilla in the west to being one of many with automakers coming into their own offering self driven taxi style service on what will probably be a subscription model (like how we have monthly payments for cars today).

There was a time, less than 20 years ago, when people would have said the same thing about Nokia.

>>Though what I do think will happen is an erosion of their lead where they go from the sole 800lb gorrilla in the west to being one of many with automakers coming into their own offering self driven taxi style service on what will probably be a subscription model (like how we have monthly payments for cars today).

Either this, or I could easily seeing aggregator apps, similar to the many that have popped up to manage multiple social media accounts, appearing to help drivers and riders manage both Uber and Lyft accounts.

They already kinda are. Look at, say, Drive Now (https://uk.drive-now.com/) - they're owned by Sixt, and they operate Drive Now jointly with BMW (BMW cars being the only type you can have with DN). I imagine automakers have relationships with other "car-sharing" firms (e.g. Zip Car) as well.

You might already know that, but for everyone who doesn't live in a city where they're available... :)

> Lyft took money from GM

GM also aquired Cruise which builds automated driving kits http://media.gm.com/media/us/en/gm/home.detail.html/content/...

Is (almost certainly) going to, but hasn't yet, acquired

Per your point that Google knows far more data than Uber via Maps/Waze, I'd like to point out that Google Ventures plowed 200mm+ into Uber at one point. So a partnership between the two is not unlikely in a future like the one your describe.

Whats the big deal? They arent decades behind their competiton. At most you can argue they are a few years behind. With their resources they can catch up quick.