Self-Driving Vehicles Are Heading Toward An AI Hurdle
If you trust the CEOs, a completely autonomous vehicle can be just months away. In 2015, Elon Musk forecast a completely autonomous Tesla by this year; and so did Google. MobileEye and Delphi’s Level 4 system is presently projected for next year, the same year Nutonomy aims to issue a number of driverless taxis on the roads of Singapore. GM will put a completely autonomous vehicle into manufacturing in 2019, with no ability for drivers to interfere or no steering wheel. There is real money behind these forecasts, bets made on the supposition that the software will be capable of catching up to the hype.
But the delusion of a completely autonomous vehicle might be further than we think. There is increasing concern amongst AI professionals that it might be years, or also decades, prior to self-driving systems can dependably prevent accidents. As self-educated systems struggle with the anarchy of the real world, experts such as NYU’s Gary Marcus are bracing for an excruciating recalibration in expectations. That delay can have disastrous effects on firms investing in self-driving tech.
On a related note, “Autonomous vehicles will hit the streets of China within next 3–5 years,” claimed the founder of Baidu (Chinese internet giant), one of the leading designers in the world for driverless vehicles, to the media in an interview.
That is a lot quicker than forecasted by information technology minister of China, who earlier claimed that it might only be reality in 8–10 years, mentioning limitations associated with security. “I am more hopeful than him, I believe it will arrive earlier,” claimed Robin Li, CEO of Baidu, claimed to the media.
Baidu, frequently dubbed as Google of China, controls the top search engine of the country and also spends a lot in services varying from payments of online to artificial intelligence and linked devices.