Turns Out, Robotaxis Are A Bit Crap
So, why are they being pushed so hard?
Self-driving AI robotaxis are meant to be our utopian future. A cheap, easy, fast, and private way to travel that could render the personal automobile a thing of the past. Well, at least, that is the lie that has been fed to us. A recent pilot study found that this technology is way off the mark and falls short of human-driven taxis on every metric. This, combined with recent revelations that AI is facing a cost and development crisis, paints a grim future for this technology. So, why are robotaxis still being so heavily pushed? Well, the answer is a little disturbing.
This study was conducted by Forbes. They compared Waymo self-driving robotaxis to ride-sharing services like UberX and Lyft over 50 different trips at different times of day in Los Angeles. Their findings were damning. On average, the Waymo vehicle’s trip time was 120% longer than the ride-share services, which is over twice as long! Not only that, these robotaxi trips also cost, on average, 30% more. So, even if you take into account tipping the ride-share driver, the Waymo was still more expensive despite massively inconveniencing the customer by taking so long.
While the Forbes study didn’t cover this factor, the quality of Waymo’s rides is widely reported to not be very good either, with allegations of dodgy driving, Waymo vehicles honking at each other repeatedly, and similarly intrusive issues. On top of all of that, despite the high price, Waymo is still miles away from profitability.
In short, Waymo is a lose-lose situation. The customer gets crap service, and the operator is losing money hand over fist.
This shouldn’t be a surprise, though. This was predicted by multiple peer-reviewed studies.
One of these studies from MIT looked to assess what percentage of tasks can be cost-effective for a business to automate with AI. They specifically looked at computer vision AI, the same technology that enables self-driving cars. After analysing over 1,000 tasks in 800 different occupations, they found that only 23% of said tasks would be cost-effective to replace with AI. Not only that, but only the most mind-numbing, completely constrained, and simple tasks could be replaced by AI, and even then, it would mostly only augment a job and only partially replace it. In fact, this is why the researchers didn’t even include robotaxis in this study, because they are such a complex and broad application of AI that it was painfully obvious that it simply wouldn’t be cost-effective.
But surely AI is going to get cheaper as it develops, solving this problem, right? Well, no.
For one, the researchers also found that even if the cost of AI declined by 20% each year, it would still be decades before it becomes cost-effective for businesses, such as taxi services, to automate themselves with AI. And the fact of the matter is AI isn’t getting cheaper. As I have covered before, AI training is beginning to hit diminishing returns, which means keeping AI development consistent will require exponentially more training, computing power, energy, and, as a result, cost. In its current guise, AI is only going to get more expensive to develop and deploy, as, for it to deliver on the automation promises fed to us by big tech, it needs to be far more capable, accurate, flexible, and robust than it currently is, which means more development.
I could end the article here. But the rabbit hole goes further down. Waymo, Tesla, and every other robotaxi company are fully aware of this problem. These issues with AI development and their questionable application in robotaxis have been well-documented in the technology and AI world for a while now. So, why are they still pushing it? They have no feasible route to profitability; their costs are only going to soar.
Well, there is a simple explanation: we are living in a post-capitalistic hellhole.
These companies know they can’t compete against taxi drivers on merit. Robotaxis will always be more expensive, and there is zero conclusive evidence that an AI driver can ever be able to drive as well as a human driver. As such, if true capitalism were at play here, these companies would simply fold.
But they aren’t competing like this; they are taking the market-share-before-profit approach so heavily pioneered by the tech world in the ‘90s.
Ultimately, as long as they can pass regulations, these robotaxi companies can completely flood the ride-hailing market. It doesn’t matter that they are worse; because their systems are automated, they can scale far faster than any human-reliant taxi service. This gets them a growing market share through brute force, not free-market forces, and in turn puts immense pressure on human-driven taxi services. At a certain point of critical mass, they can price gouge the customer to reach profitability, as there are not enough human-driven taxis to give them enough competition and too many robotaxis for there to be an incentive to start any new human-driven taxi services.
They aren’t competing on merit; they are competing on volume, and we will be the ones to pay the price for that, as we will end up paying more for a crappier product.
This prediction isn’t unfounded or hyperbole. Investors in self-driving taxis are constantly talking about market share before profitability and how AI being unable to reduce itself in cost isn’t an issue. In fact, those pouring millions or even billions of dollars into these technologies explicitly talk about how this service can manipulate the market in their favour, not compete fairly within it.
So, next time you see a Waymo or hear Musk babble on about his non-existent robotaxi, take a moment to ask why they are pushing this questionable technology onto us.
Thanks for reading! Content like this doesn’t happen without your support. So, if you want to see more like this, don’t forget to Subscribe and help get the word out by hitting the share button below.
Sources: Futurism, Futurism, Forbes, Will Lockett, Will Lockett, Will Lockett