I work in the auto industry. First at GM in manufacturing an then with 3 tier 1 suppliers. I follow technology for a living as an innovation manager for the company I work for. Here's my beef: Tech companies are good at creating a vision of a future and then repeating it until they get others to believe it. There timing is always woefully overstated. Then you have one of the most competitive markets in the world start to hedge their current business by investing massive amounts into R&D to try to ensure the threat from all the new entrants won't put them out of business. There will be a hundreds of billions of dollars wasted on this effort. And, many of these companies can't afford the investment dollars, but can't afford not to make the investment. Trust me, the next market downturn will see car companies go bankrupt again.

So what needs to be solved to actually move to autonomous vehicles:
-Laws. Governments around the world need to define who is liable for what when something goes wrong and someone dies. The conversation has barely begun.
-Actual technology that can be implemented into mass produced vehicles. As mentioned, LiDAR is large and completely out of line with cost requirements. First, look at the picture of the LiDAR system on top of the car shown in this article. Do you think aerodynamics have a place in a functional car? Do you think you would have some wind noise issues, some high speed induced vibrations? Do you think an automobile needs to efficiently move it's weight with least drag forces? LiDAR needs to shrink by a factor of 100. And the cost needs to scale accordingly.
-Automobiles are one of the most costly consumer products and one of the worlds most competitive markets. It's really hard to design and develop and mass produce automotive components and vehicles. The vehicle costs cannot be jacked up $7,000-$10,000 per vehicle. Automobiles are price elastic, meaning if you raise their price, in general, the demand craters. I believe it was in S. Korea where the local government incentives for electric vehicles came off recently. The sales crashed to almost zero.
-There is an assumption that is made but rarely talked about. Autonomous vehicles assume they will be electric vehicles or perhaps fuel cell. There needs to be a massive leap in battery technology and charging to enable range and charge time to be radically improved. And there needs to be massive capacity to build batteries for all the proposed cars. Last I read, there are well over 100 electric vehicles that have been announced. Good luck with charging infrastructure and batteries capacity.
-Software and electrical systems need to be secure from the rotten people of this world intent on doing bad things.

So in summary, we may have technology that can function today for an autonomous vehicle. But we are years away from solving these massive issues. I was at CES this year and attended the Automotive conference sessions. I was shocked by the arrogance of the tech companies. I most shocked by NVDA. Someone asked a question and the executive said: "People like to talk about ethical dilemmas in autonomous vehicles. They don't exist. We will never have to make a choice about hitting a tree or hitting a child. Our technology sees 360 degrees. So we will never be in that situation." Ummm. I'm quite sure he will be proven wrong. The technology will always need to make these decisions. But he got points for his effort on trying to manipulate the market...