AI-controlled cars were touted as a panacea to our transportation woes. But are self-driving vehicles really the answer?
Comment
Autonomous cars are the future, they say.
Safer, smarter time savers, they say.
A paradigm shift for civilization, they say.
Well, guess what? Artificial intelligence-controlled autonomous cars, robotaxis, self-driving vehicles, or whatever else you want to call them have a long, long way to go before they change our lives — if ever.
I’m not an autonomous vehicle (AV) hater. The idea of a part-time machine that allows you to roll out of bed and hop into your driverless robotaxis and then wake up at your destination after three hours of complete rest is a beautiful idea.
But spare a thought for residents of San Francisco (the global epicenter of AV testing due to California laws and proximity to Silicon Valley), who routinely face lengthy holdups due to malfunctioning AVs as part of extensive robotaxi trials.
or worse.
According to AVCrashes.net, there have been 623 autonomous vehicle crashes in the greater San Francisco Bay Area since 2014, and 18 related deaths nationwide as of 22 Jan 15, 13, according to the USA’s NHTSA (National Highway Traffic Safety Administration).
Getty Images
OK, so we don’t see a Code Red situation (Skynet, anyone?) and robotaxis are extremely useful for many, but San Fran residents regularly complain about the traffic snarls caused by grinding AVs because they can’t. We deal with endless situations we see on the road every day.
Reports of AVs impeding public transportation are common, and the San Francisco fire chief has reported nearly 40 incidents in which AVs have ‘interfered’ with emergency responders.
Waymo and Cruise, the two largest robotaxi operators in the San Francisco Bay Area (owned by Google and General Motors, respectively) are expanding their fleets of self-driving vehicles, and some are now engaging in street wars with AVs.
Vigilante groups have started a battle with AI-controlled vehicles by putting traffic cones on their bonnets, causing ridiculously complex and expensive trains to be confused and canceled.
According to The Guardian, various activist groups, including Safe Street Rebels, are mobilizing residents to take action against San Fran’s automatons.
Even taxi and Uber drivers aren’t too fond of robotaxis (to put it mildly!) because their revenue lines the pockets of the obscenely rich and powerful tech and automotive companies.
AV advocates have touted the technology as a silver bullet for road injuries, as 90 percent of road deaths are caused by human error and point to the extra leisure time it could afford us all, but a leading Australian university, the Royal Melbourne Institute of Technology (RMIT), has urged caution. .
Already-common Level 1 and 2 automated driving systems such as autonomous emergency braking, active cruise control and, at least for short periods of time, lane-keeping make life behind the wheel easier, if decidedly streamlined.
But fully hands-and-feet-off Level 3 and 4 autonomous vehicles – which must complete Level 5 driverless cars before blindfolded cars like the ones being trialled in San Fran become legal on a wider basis – are not necessarily safer than conventional cars. .
Dr Neng Zhang from the RMIT School of Engineering and lead author of the study “Does driving experience matter? It examined drivers’ takeover performance in conditional automated driving and how experience and various distractions such as work, social media, and leisure affect the driver’s ability to respond during an emergency.
“We asked them to write business emails, watch videos and rest in a resting position with their eyes closed,” said Zhang of Level 3 AV Simulation.
“We found that rest produced the worst takeover response, followed by work. Social media was less disruptive. However, the longer the participant engaged in an activity, the worse their response to the emergency.”
Zheng’s colleague and co-author Professor Mohammed Fard explained the need for federal and state governments to ensure that self-driving cars do not become social disasters.
“Governments can effectively protect road safety by identifying these harmful effects and regulating non-driving activities in the context of autonomous driving,” the professor said.
AI is also facing an image crisis in Australia. And no, ChatGPT didn’t write this article and won’t write anything with my name beyond pointless insults to colleagues going abroad to ‘find themselves’.
Indeed, in a Roy Morgan survey of over 1400 Australians aged 16 and over, co-funded by the Campaign for AI Safety, 57 per cent of Australians believe AI ’causes more problems than it should’.
Should we be worried that autonomous vehicles could one day develop manipulator weapons and enslave us all? It’s unlikely, but a Roy Morgan poll found that one in five Australians think ‘AI threatens to wipe out humans in the next twenty years’.
Even if AVs manage to spread globally – once legal issues of liability in the event of a collision are resolved – they’re unlikely to completely displace manually driven vehicles, but even BMW doesn’t think self-driving cars will go global.
According to Volkswagen, AVs will be mainstream by 2030 but whether that includes countries like Australia, or places with poor lane markings and road signs, remains to be seen. A former Holden technical engineer reckons there are plenty of obstacles – particularly Australia’s unique road signs.
AVs becoming mainstream sounds like peak oil or the Y2K bug to me.
Fast forward two decades and there was a far less cynic Feann Torr in Germany, speaking to a senior engineer at one of the Big Three luxury marques who insisted fully autonomous cars would be commonplace by 2020.
Of course they are not and never will be. In fact, I think autonomous flying cars may have a good chance of becoming more ‘common’ sooner rather than later. Therefore, in my opinion, fully autonomous vehicles are a crock of sh*t.