Given that development of a new car could take between 5-10 years I think it makes more sense for Apple to start out fresh - it will cost a lot less and they will able build exactly what they want.
That timeframe is for an established company. For someone operating in a completely foreign space could be even more significant, not to mention costly. Keep in mind that car companies have spent well over $5 billion just to produce a redesigned model of an existing vehicle.
This is why I would invest heavily in Tesla if I had money that I could afford to invest.
I'm too skittish to pull the trigger on TSLA. Understandably, they a new company trying to enter a market with high barriers so costs are high. However, they have set their goal to make a profit in 2020. Last year they lost about 300 million on 3 billion in sales. The value in Telsa is largely through their stock, and the expectation that they will do great things in the future.
There will be no autonomous vehicle on the road until the highway dept. and the federal government upgrades all highways and I don't see congress doing that.
The technology for self-driving cars is already sufficient to replace human drivers and reduce total number of accidents. You are correct though in that it will be slow progress to allowing that. This is why you see so many vehicles introducing safety technology that "takes over" when you're about to do something stupid. It's a way to familiarize drivers with the root technology. Humans have an illogical aversion to letting technology replace them as drivers. As a race, we are horrible pilots.
Car makers are increasingly using touch screen interfaces
In addition to voice systems...
Of course they have the money, that was never my point... spending billions when you have billions to spend is the easy bit.
Well, it's a little bit harder too. Apple holds the vast majority of their cash overseas to avoid paying taxes on it. If they were to use it for a purchase, it would have to come "home" for taxation first effectively increasing the cost of such a spend.
But, Apple doesn't need to sell that much to be a success. If they sold 2-3M cars a year within 8 years, that's be a fantastic success and the fastest launch of car brand ever.
It would be impressive to sell 3 million cars in 8 years. It's also an improbable suggestion. You're talking about them controlling 20% of the US auto market in their first 8 years? I don't think any company would even dare suggest such a thing was possible. Tesla wants to sell 55,000 this year, and they are the golden boy of the industry. There's so many different reasons that the mind reels from the idea. Apple isn't a manufacturer, and they don't have a significant retail presence. Where will these vehicles be sold? Who will make them?
Tesla is the only serious electric car company at this point
The only serious electric-only car company...maybe.
My best bet is they will take an existing car model -like Toyota Corolla- modify it and then call it their cars (Like the A8 chip?).
You're conflating dissimilar industries. In the consumer electronics space it is common to be a "parts assembler" for the end user (which Apple isn't). It has become incredibly more common in the automotive industry to be "fancy" parts assemblers, but they're still doing a huge amount of heavy manufacturing. Your suggestion with the chip would be analogous to the car's engine. I see no reason to believe why Apple would be better at creating engines than the people out there.
Actually, I think the chances are greater that Apple releases a line of TVs before entering a cutthroat business like the auto industry.
While I'd quibble with the suggestion that television sales aren't cutthroat, you're spot on here. This is a consumer electronics market where suppliers are readily at hand, and you can pay someone to be your assembler. This is where Apple knows how to play.
The partnership with IBM may hint some great opportunities to develop technical solutions to manufacturing issues that may help car companies to solve plus enable harvesting data from vehicles in a away that will help them capture actual customer vehicle usage for future Apple Car solutions.
Everything I've heard from IBM about the deal has been much more pedestrian. It feels like Apple wanted business to take them more seriously, and IBM wanted people to think they're cool with no amazing plans. Truthfully, toolkits and core products to allow people to make "smarter" offerings and leveraging IBM's technology with Apple's name and ecosystem.