"Once self-driving cars hit, the switch will be immediate and massive. I would predict 10% to 90% penetration in probably less than 12 months--the limitation will simply be building the cars fast enough. That's a lot of people suddenly with no income."

It will probably take 12 months just to form the exploratory committee that will form the actual committee that will even begin to consider legalizing fully driverless cars.

This change involves many levels of government bureaucracy. It will happen, but it's definitely not going to be "immediate".

Depends on the government. Beijing could go driverless overnight if Xi Jinping thinks it's a good idea. And considering Beijing's extreme traffic woes, they just might....

> It will happen, but it's definitely not going to be "immediate".

Normally, you would be correct. However, there will be so much at stake that there will be huge amounts of money greasing the skids on any required legislation.

Real Level 3 Autonomy will displace 75% of truckers almost immediately--that's 1.5 million jobs lost. You will have one or two drivers in the front truck and a half-dozen trucks running autonomously but slaved to the front truck. WalMart will almost certainly lead the charge.

Level 4 autonomy will allow the insurance companies to jack insurance rates on normal drivers through the roof. You will get a massive insurance discount if you never touch the wheel (removes human driver error--total monitoring so always other driver's fault).

With that much money at stake, the obstacles will melt like ice in the summertime. See: Fracking in Pennsylvania for an example of how fast legislation can move with enough money behind it.

Fracking, for all its controversy, does not outright kill random people when something goes wrong. Whenever there's a problem, it's going to be front page news, and people (old people in particular) are going to call their political representatives and complain. This is going to become a massive political football. There's no way it gets introduced smoothly and quickly--that's just a futurist daydream.

No that's not how auto insurance works. It's a commodity with high competition and low profit margins. Insurance companies set prices based on expected losses. There's no reason to think that the arrival of level 4 autonomous vehicles would cause an increase in losses on human driven vehicles. If one insurer jacks up prices then the others will undercut them to gain market share.

The "once self-driving cars hit" point would be their legalization in your scenario, with the same rapid uptake described unless government implements some sort of quota system allowing only a small number to be added at once.

I'll buy that. My point is that it's not something that will happen all of a sudden. First, the tech needs to be ready. Then we'll have a years-long period of political wrangling. And once we finally get through that, sure, adoption could be swift. But people in tech who think this is just going to sail through aren't in touch with how the majority of people look at the world... they're sort of like the folks who thought Google Glass would be a smashing success.

> But people in tech who think this is just going to sail through aren't in touch with how the majority of people look at the world... they're sort of like the folks who thought Google Glass would be a smashing success.

Maybe, maybe not. We've seen tech outpace regulation many times - Facebook and privacy, Uber and taxis, e-mail and spam, etc.

This process is already under way. There are hearings in Congress regarding autonomous vehicles several times a month.

If we had less rules, we could make it happen. More industry standard regulation, then things could evolve much more naturally and not have to be lobbied for all the time.

So it's a bad idea that we're thinking about what we're going to do before doing it?

I mean, I can understand "move fast and break things" when it comes to releasing web applications. My sympathy ends when you're literally moving fast and literally breaking things.

12 months? Perhaps for taxis and private hire vehicles, but not for the population as a whole. In the first year, perhaps 10% penetration, but more likely 2-5%, as for most it will be a premium feature. I'd say 5-10 years for normal models to get full self driving capability.

People are also it's beneficiaries

Problem is, the beneficiaries and the victims are often different people.

If they were the same people, the technology would be rather pointless.