By 2035 there will be 76 million autonomous cars on the world's roads, this according to IHS Automotive. The company made its prediction based on recent advances in tech as well as the numerous partnerships recently announced between ride-sharing companies and automakers. Earlier this year GM invested $500 million in Lyft with the end goal being development of a fleet of self-driving taxis. Uber, Lyft's main competitor, partnered with Toyota. And then there are the automakers going it alone, like Jaguar and (maybe) Ford.

And amid all those partnerships is the continuing work that Google is doing on its self-driving car and the potential driverless car Apple is said to be cooking up. Hell, just about every major automaker has an autonomous car project or two on the books. The IHS believes the US will be an early leader in self-driving car deployment but that Japan will help speed things up as the 2020 Olympics in Tokyo arrive. From there the forecast is 600,000 global sales of autonomous cars in the year 2025, with the US alone expected to sell 4.5 million autonomous cars per year by 2035. However, this number is dwarfed by the 5.7 million self-driving cars predicted to be sold in China that same year.

But still, 76 million is a lot of autonomous cars to sell in less than 20 years. While everyone would love to own a self-driving car, the reality is that new automotive tech is often buggy. Normally these annoying tech faults can be ignored, such as a navigation system that doesn't work. However, if the tech controlling an autonomous car starts to crap out then the entire car is rendered useless. Another thing that could prove this forecast ambitious is government red tape. Numerous states have shown that they can't even regulate companies like Uber and Lyft, so how the hell are they (or the federal government) supposed to set rules and laws for self-driving cars?

Finally, new tech in cars is often expensive. This high cost of entry means that autonomous cars may not exactly fly off dealer lots, at least not initially. It's one thing to sell a top-tier Volvo XC90 that drives itself. But how about a Toyota Corolla with the same tech? Yes, autonomous cars are coming but 21 million by 2035 seems like a hefty overestimate.