Beware the Credit Crash!

kabayiri

grawiz wrote: »

If you can hold on until Brexit actually starts, there should be some amazing deals available as the economy plummets.

It's a weird effect.

In a sales downturn, not only do new car deals improve, but quality used cars gain in value as people hold on to their cars for longer.

GreatApe

WengerIn wrote: »

I do wonder about self driving cars.

Firstly there's the problem about how we get there. I'm not sure I want a road full of AVs and people trying to game claims against their insurance. Secondly I'm not sure I want to sit in an AV that is going to return control to me when disaster strikes. Thirdly I don't want to be doing 80 mph and get a phone call asking me to pay £20k to return braking control to me.

I love the idea, I think most people do. The reality might be a harder sell.

Plus vehicle number savings might be lower than you think. You claim 2/3rds. Well I don't know about you but my kids have to be at school at 9am and I need to be in at 8.30am as does my wife at a different place of work. If my kids are going to be taken to school by an AV then I need more cars not fewer!

With AVs the problem remains that most people work something approximating 9-5, that's why we have a rush hour. Like with power stations we need enough capacity for peak hours not average consumption.

Self drive is just an information processing problem. Its only a matter of time for computer processors to get to the point where it is possible to process the information better than a human can.

So the argument is only one of time not one of possibility. I think 2020-2025 seems about the time we will have the hardware and software but you might think it will be 2030-2035 but it is coming.

Computers will not be 100% safe but nor are humans. Computers only need to be better than humans which will definitely happen in time because computers can use senses we can not. So they will use vision like we do when we drive but they will have a dozen eyes not two they will have radar and ultrasound and possibly 3d laser mapping. It is inconceivable that a human will a jelly brain and two eyes will best computers for much longer.


Yes it is true that traffic has peak periods however personal cars are so large in number that its never ever close to 100% of cars on the road at the same time. As an example earlier this year my household has already gone from 2 cars to 1 car + uber. Just the savings on insurance for the second car will probably be cheaper than the uber trips and the savings in fuel depreciations maintiance etc is mine to keep. Self drive cars will just be a much cheaper version of uber.

As for your issues of hacking I think that's going to be much less of a problem than other human drivers that are drunk or high or just stupid crashing into you. Likewise the risk of you passing out or having a heart attack while driving or falling asleep at the wheel while low isn't zero. The risk of hacking will be lower than the risk of a driver falling asleep.

Also there is likely to be a pool function. So cars with 4-6 passengers. Eg like your kids going to school a self drive taxi will pick 6 of them who live nearby and take them to the school and drop them off. So 1 car does the job of 6 cars + parents. You might think sharing a car is unsafe or something you don't want to so but millions of people share a bus and tubes and trains these self drive cars will be more secure as they can have better CCTV and also would keep a record of who has booked them plus they can be designed to be compartmentalised so the passengers all have their own bit of space.

It will also kill trains. A self drive taxi seating 9 passengers doing a non stop Birmingham to London trip will be faster than trains (no need to stop at half a dozen stations) and can be point to point. The fuel cost of such a trip would only be £3 the capital cost about £10 and the maintenance insurance etc probably less than the fuel cost. You have a cost base in the region of £2 per passenger. Short distance intercity trains will die. Their tracks either abandoned or converted to roads.

westernpromise

What sort of car does 125 miles on £3 of fuel? That's 227 miles per gallon. Did you mean £30?

The issue I would have is that Tony Blair's version of the software would be different to mine and might even control my car, so that in the event of a collision it would be me rather than him who got killed.

GreatApe

antrobus wrote: »

I believe that the GreatApe is somewhat over-enthusiastic over the impact of self-drive cars.

As you have noted, there is the issue of peak demand. But more importantly, as you have again noted, there is the open quesion of whether people want the darn things. We already have taxis, people can already decide to ditch the car on their driveway and rely on taxis, but they choose not to do so.

So it's not a question of arithmetic, as in x number of self-drive taxis can replace y number of owned cars, it's a question of economics, as in dispensing with the need for a human taxi driver will make taxis so much cheaper that people will make a different choice.

Well yes over 500,000 miles a human computer wants £250,000 while a silicon computer will cost less than £5,000 and that will further drop to £500 in a decade or two.

The lower prices will drive people to using more computer taxis and many households will drop their second vehicles which will mean even more people use computer taxis which will make them cheaper and quicker which will drive more people to using computer taxis. For a 16 year old the up front costs of getting driving lessons and paying £2k insurance and buying a crap second hand car night amount to £5k in total. Many young people will just not bother if computer taxis cost 25p a mile why pay £5k upfront + 15p a mile for fuel/maintiance? Why would parents pay for driving lessons when they know their kid driving a car is 100x more likely to kill or harm themselves compared to using a robo taxi.

But beyond the huge savings of not needing to pay a human computer £250,000 and getting a silicon computer for £500 there will be large savings in vehicle capital costs. Taxis drive for 500,000 miles while personal cars more typically do 100,000 miles in their life so the cost per mile for the car itself falls by 80%. The capital cost of vehicles is often more than the fuel or maintiance so a 80% saving on capital cost is huge. Insurance will be lower too even if the risk and crashes are the same the fact that fleets can buy insurance in bulk will mean cheaper deals.

I think its going to be here soon by 2030 half of all miles will be computer driven not human

westernpromise

Why won't JohnnyCabs cost the same as ordinary cabs? If a journey costs £10 by minicab, the JohnnyCab's going to charge you £9.50, isn't it? Or indeed £20, if they're so great?

Value =/= cost. Just because it can be done for £3 doesn't mean it will be sold for £3. If it's better than a normal cab the charge will be higher, not lower.

antrobus

GreatApe wrote: »

...I think its going to be here soon by 2030 half of all miles will be computer driven not human

Your opinion is noted.

Your prediction of future taxi prices might be right, or it might be wrong. Or it might be somewhere in between.

But I don't think that it will impact on the prospect of the 'threat' of PCP driven credit crash in 2017.

GreatApe

westernpromise wrote: »

What sort of car does 125 miles on £3 of fuel? That's 227 miles per gallon. Did you mean £30?

The issue I would have is that Tony Blair's version of the software would be different to mine and might even control my car, so that in the event of a collision it would be me rather than him who got killed.

125 miles at 4 miles per kWh = 31.25 kWh of electricity x 10p per unit = £3.12 for a 125 mile trip. That is possibly an over estimate as fleet operations can probably buy electricity cheaper in bulk than you or I can for our homes especially of they charge up at off peak hours or at night.

Safety will be better than humans. Electric cars can also be designed to be safer as the front does not have a large solid metal engine so they can have a larger crumple zone. The human body can survive extreme deceleration. The USA air force did lots of experiments in this forcing some people to decelerate at over 100g (which is like going from 100mph to zero over just 0.05 seconds) it blinds people temporarily but vision returns. Without a steering wheel so close and the larger crumple zones and the fact that cars will less frequently get into crashes safety will be better for all. I doubt there will be classes of safety where you pay more for more safety that would be stupid. Even if that were the case which it won't be (regualtora wouldn't stand for it) as long as the worst safety level is better than a good human driver then ota still better than now

antrobus

westernpromise wrote: »

What sort of car does 125 miles on £3 of fuel? That's 227 miles per gallon. Did you mean £30?...

An electric one? The Energy Savings Trust states that "a charge giving you a range of around 100 miles will cost around £2 to £4 in electricity".

http://www.energysavingtrust.org.uk/travel/electric-vehicles

I have no idea as to whether their calculations are correct.

westernpromise wrote: »

..The issue I would have is that Tony Blair's version of the software would be different to mine and might even control my car, so that in the event of a collision it would be me rather than him who got killed.

Since we have yet to see any self-drive cars, it is not known what kind of software will be available. Or indeed, whether or not the available software will be exactly what people want.

GreatApe

westernpromise wrote: »

Why won't JohnnyCabs cost the same as ordinary cabs? If a journey costs £10 by minicab, the JohnnyCab's going to charge you £9.50, isn't it? Or indeed £20, if they're so great?

Value =/= cost. Just because it can be done for £3 doesn't mean it will be sold for £3. If it's better than a normal cab the charge will be higher, not lower.

Competition will mean these self drive computer fleets can not charge monopoly prices. The companies that get there first will have huge profits but the savings even with the profits will be massive. For instance in the USA alone its predicted to be 6 trillion miles by vehicles in 2030. If the fleets charge 3 cents a mile (the human taxi driver charges over 30 cents a mile) that is $180 billion annual revenue with margins in excess of 50% as its just software. A company that had p/e of 15 x would be worth north of $1.5 trillion. And that is just one country. Multiply by 20 x to get world milage and you can see the huge profits to be made by the companies that bring to market a working competitive self drive vehicle.

Even at 0.5 cent per mile x 120 trillion miles done in the world = $600 billion revenue and since its software it will be very high margin. Its likely to be an American company to bring this to market and they tend to opt to do zero margin until they have got a lot of the market share think amazon or uber operating at close to zero or even negative margins for years yo try and grab market share.

westernpromise

antrobus wrote: »

An electric one? The Energy Savings Trust states that "a charge giving you a range of around 100 miles will cost around £2 to £4 in electricity".

http://www.energysavingtrust.org.uk/travel/electric-vehicles

I have no idea as to whether their calculations are correct.




Since we have yet to see any self-drive cars, it is not known what kind of software will be available. Or indeed, whether or not the available software will be exactly what people want.

I take any promises from the green blob with a large pinch of salt. Look how well they did on diesel.

I also find it very unlikely that the PM, or Tony Blair, would be driven around in a car with exactly the same software as every prole's car. There would certainly be VIP versions of the code that killed everyone else in an accident first. Oiks like us would not be able to buy it, grandees such as Blair and Corbyn would get a special version for nothing.