Electric car confusion

Autoblog titles: The race for autonomous cars is over. Silicon Valley lost . The point they want to make is:
To paraphrase Elon Musk, Silicon Valley is learning that “Making rockets is hard, but making cars is really hard.” People outside of the auto industry tend to have a shallow understanding of how complex the business really is. They think all you have to do is design a car and start making it. But most startups never make it past the concept car stage because the move to mass production proves too daunting.
and
Yet, while companies like Google and Apple are giving up on making cars, they’re not giving up on the auto industry. There is another area where Silicon Valley could play a dominant role and it’s all about accessing car-based data.
It’s about the margins - making a thing gives you around 10% markup, making things out of data gives you much, much higher margins. This also frames the current discussions about privacy within the German government, and who is to own your data (hint: not you), especially when you drive.
But making software systems is also really, really hard. And an electric, self-driving car is much more a system, then a piece of software and then a piece of hardware. And Forbes sees the automotive industry not really in a strong position here, focusing on GM and Tesla.
In reality, I don’t see how GM isn’t making a meaningful strategic pivot to electric, especially compared to Tesla. Why? By March 2017, GM sold a little over 2,000 Bolts, which retail around $37,000. By way of comparison, Tesla has delivered “over 183,000 Model S and X vehicles over the last four years,” according to Electrek. And then there’s the Model 3. Tesla has received over 400,000 customer deposits for its Model 3, which is priced at $35,000.
What GM does with the Ampera-e looks like a compliance stunt to get fleet consumption down, not like a credible electric move, to some: Greenwashing?
So what to make of it? Making cars is hard. Making self-driving, electric cars is even harder, because it involves the hard tasks from making large physical machines and the hard tasks from making distributed, secure software systems. Traditional car makers are not delivering (GM singled out above, but the others are guilty, too). And are not pivoting aggressively and fast enough into software and into electric. On the other hand, Silicon valley has largely pulled out of making large physical objects, and focuses on the data generated by drivers. Only Tesla seems to move at speed.