By Lance Eliot, the AI Trends Insider
Some suggest that the makers of AI autonomous cars ought to share openly with each other and freely reveal their proprietary wares, doing so to enable heightened progress in achieving true self-driving cars and presumably therefore obviating the 40,000 car crash related deaths each year in the United States alone and an estimated 1.3 million or more fatalities that occur worldwide annually via car accidents.
The sooner autonomous cars are put onto our roads, the sooner it is said that we will start saving precious lives and reducing car crash induced injuries.
In a kind of presumed kumbaya moment, akin to at wartime having everyone suddenly lay down their arms, wouldn’t it be wonderful if the automakers and self-driving tech firms would divulge everything that they know and are arduously doing toward creating a self-driving car, and then the resulting synergy of this massive “open source” approach might speed-up the process?
The point being that today there is an existent winner-beats-all mindset, of which each automaker is striving on their own to craft a true autonomous car, but wouldn’t we all be winners as a society if the automakers just opened-up the kimono and bared all for everyone else to see their self-driving car efforts.
Presumably, they ought to showcase the proprietary algorithms that they’ve devised for self-driving capabilities and post their voluminous AI code in a public forum for all to peruse.
They ought to release the data that they’ve collected from their roadway trials, providing a treasure trove of data that could allow others to employ Machine Learning and Deep Learning to ferret out ways to best drive a car.
For those using simulations to do testing of their AI driving systems, they ought to allow others to log into the simulation and see what kinds of parameters and settings are being used, along with allowing anyone else to utilize the simulation for the furtherance of their driverless car tech.
Furthermore, one of the biggest hurdles in devising self-driving cars involves figuring out so-called edge cases, involving seemingly unusual or oddball driving circumstances and being able to prepare the AI driving system to cope with those situations.
Rather than each automaker having to figure out edge cases on their own, perhaps it would be best if a large-scale database of known edge cases was formulated and provided access to anyone interested. Think of how such a collective set would aid others that hadn’t yet discovered various edge cases and thus they would not have to reinvent the wheel, so to speak, and could simply refer to the open database instead.
In short, the assumption being that today’s rather fragmented and disjointed approach involving singular companies attempting to each develop AI-based self-driving cars could be turned on its head, allowing a grand collective of all such makers that would open their doors and share their proprietary secrets.
Sure, there are some automakers that have already done something like this on a smaller scale, forming partnerships or joining small consortiums of fellow firms, but the notion here is a wholescale and unmitigated sharing and openness by everyone in the industry.
In theory, this would also encompass the people involved too, namely that the AI developers, engineers, scientists, and others would toss aside their company badges and become a member of a national or presumably global collective seeking to produce true self-driving cars.
Many have likened trying to achieve self-driving cars to be like a moonshot, and thus you might suggest that this desire to get everyone to share is akin to asserting that the mission to the moon or similar is so important to humanity that all should reveal their efforts.
Imagine a science fiction movie whereby the earth is under siege and our only hope is to colonize and live on the moon. As a species, we would want to collectively work together to find a means to lift all humans to the moon and set up moon-based cities for us to live there.
If we don’t all pitch in toward the moon mission, it would mean that some humans or perhaps all humans would be wiped out once the threat to earth actually happens.
I’m not suggesting that the lack of achieving self-driving cars is on par with such a cataclysmic saga, and only trying to offer that it is roughly the same kind of logic being employed.
In short, the propositional argument is that the automakers ought to make fully publicly known their self-driving wares since doing so would presumably save lives by sooner achieving driverless cars, based on the assumption that other such firms could all leverage the heretofore secreted and isolated efforts.
We would all use each other’s work as stepping stones, climbing up the steep ladder toward driverless cars, together.
Today’s question then is: Would it be viable for all automakers and self-driving car tech firms to share everything they’ve got in order to presumably speed-up the pace of arriving at true self-driving cars?
For numerous reasons, as I’ll argue herein, it isn’t a viable proposition.
Let’s unpack the matter and see why.
For my framework about AI autonomous cars, see the link here: https://aitrends.com/ai-insider/framework-ai-self-driving-driverless-cars-big-picture/
Why this is a moonshot effort, see my explanation here: https://aitrends.com/ai-insider/self-driving-car-mother-ai-projects-moonshot/
For more about the levels as a type of Richter scale, see my discussion here: https://aitrends.com/ai-insider/richter-scale-levels-self-driving-cars/
For the argument about bifurcating the levels, see my explanation here: https://aitrends.com/ai-insider/reframing-ai-levels-for-self-driving-cars-bifurcation-of-autonomy/
The Levels Of Self-Driving Cars
It is important to clarify what I mean when referring to true self-driving cars.
True self-driving cars are ones where the AI drives the car entirely on its own and there isn’t any human assistance during the driving task.
These driverless vehicles are considered a Level 4 and Level 5, while a car that requires a human driver to co-share the driving effort is usually considered at a Level 2 or Level 3. The cars that co-share the driving task are described as being semi-autonomous, and typically contain a variety of automated add-on’s that are referred to as ADAS (Advanced Driver-Assistance Systems).
There is not yet a true self-driving car at Level 5, and we don’t yet even know if this will be possible to achieve, and nor how long it will take to get there.
Meanwhile, the Level 4 efforts are gradually trying to get some traction by undergoing very narrow and selective public roadway trials, though there is controversy over whether this testing should be allowed per se (we are all life-or-death guinea pigs in an experiment taking place on our highways and byways, some point out).
Since semi-autonomous cars require a human driver, the adoption of those types of cars won’t be markedly different than driving conventional vehicles, so there’s not much new per se to cover about them on this topic (though, as you’ll see in a moment, the points next made are generally applicable).
For semi-autonomous cars, it is important that the public be forewarned about a disturbing aspect that’s been arising lately, namely that in spite of those human drivers that keep posting videos of themselves falling asleep at the wheel of a Level 2 or Level 3 car, we all need to avoid being misled into believing that the driver can take away their attention from the driving task while driving a semi-autonomous car.
You are the responsible party for the driving actions of the vehicle, regardless of how much automation might be tossed into a Level 2 or Level 3.
Self-Driving Cars And Pace Of Development
For Level 4 and Level 5 true self-driving vehicles, there won’t be a human driver involved in the driving task.
All occupants will be passengers.
The AI is doing the driving.
As might be apparent by reading the headlines of the news, promises of being able to achieve Level 4 and Level 5 have been made and then broken, repeatedly, and some are beginning to wonder whether the any-day-now predictions hold any water.
As such, those that proffer the question about whether to open the kimono are sincerely grasping for a means to get things further prodded along.
Consider the reverse side of the question.
Why would automakers want to openly share all of their hard-fought-for and expensively devised proprietary tech about driverless cars?
Think of it this way.
Shareholders of automakers and self-driving car tech firms have bought those stocks under the belief that it is a sensible investment. If you invested in automaker X, and automaker X is able to achieve driverless cars, the stock that you hold would presumably rise in valuation as a kind of reward for the automaker having arrived at such a goal.
And, the automakers themselves have likely already invested millions upon millions of dollars, in some cases billions, toward their self-driving car development efforts.
If they laid bare all their proprietary efforts, it essentially would mean that they have tossed away their investments, and it means that other firms that didn’t make such investments are now richer for it.
Yes, maybe this would be for the collective good, but it would pretty much undermine and even wipe out the firms that long fought to try and make driverless cars.
The shareholders would likely lose their shirts, the automakers would see a precipitous drop in their valuations, and you’d be rewarding others that hadn’t devoted time and money in the same manner.
In fact, a strong case could be made of malfeasance toward the management teams and operating officers, and boards of directors, due to simply giving away their valuable assets.
So, the first argument in opposition to the open kimono notion is that there’s really little or no financial incentive for the automakers to do so, and likely a humongous financial penalty if they did so.
Next, let’s consider whether the collective approach is really going to get you what you want.
In other words, the base assumption is that by collectively sharing everything known about AI self-driving cars, we would be better off and sooner able to achieve driverless cars.
That’s not necessarily the case.
Right now, there are Darwinian juices spurring each of the automakers toward the goal of achieving true self-driving cars.
There are billions of dollars to be made, perhaps trillions of dollars, and the pot of gold is enabling automakers to spend their precious time and money accordingly.
If all automakers were going to at all times bare all on their driverless car efforts, what is the incentive for them to continue to strive mightily in that direction?
Suggesting that they ought to want to save lives is a nice idea, and certainly they would all agree that is indeed vital, but if it suggests that none of them will make money to remain in business or they will make smaller money due to a splintering of the driverless pie, you are knocking out a key motivation that keeps the engines running on achieving self-driving cars.
And, here’s a crucial point, having everything on the table about self-driving cars is not necessarily going to speed-up development.
It could have the opposite effect.
There might be so much to choose from, and the myriad of approaches so diverse, you could end up with a breadth approach rather than a focused approach, and slow down development accordingly.
Too much of a good thing, one might say.
Think too of the complexities involved in somehow coordinating across all of the industry.
Who decides that these set of sensors are the “best” approach for achieving self-driving cars, or that algorithms N and M are the best choices?
You could say that just let whoever do whatever they want, but this would seem to be a muddled way to proceed, and it would be less likely to go as fast as would the existing focused efforts.
Some might counter-argue that a special conglomerate could be established and become the centralized overseer of the aims to achieve self-driving cars.
Maybe each automaker would get a share of the conglomerate as based on the value of the driverless tech that they contribute to the collective.
Well, I’ll just say that this would become a huge drain of time and attention, while everyone fights for their relative apportionment, and the odds are it would distract from the efforts underway to arrive at true self-driving cars.
One would also have to wonder whether this special conglomerate might get bogged down and by its own weight become a barrier to advancing self-driving cars.
For why remote piloting or operating of self-driving cars is generally eschewed, see my explanation here: https://aitrends.com/ai-insider/remote-piloting-is-a-self-driving-car-crutch/
To be wary of fake news about self-driving cars, see my tips here: https://aitrends.com/ai-insider/ai-fake-news-about-self-driving-cars/
The ethical implications of AI driving systems are significant, see my indication here: https://aitrends.com/selfdrivingcars/ethically-ambiguous-self-driving-cars/
Be aware of the pitfalls of normalization of deviance when it comes to self-driving cars, here’s my call to arms: https://aitrends.com/ai-insider/normalization-of-deviance-endangers-ai-self-driving-cars/
Let’s consider some additional reasons why the grand sharing approach is not especially viable.
If sharing is so compelling and it has to do with saving lives, one must ask why the same question doesn’t get pressed in other domains.
For example, there are about 635,000 deaths each year in the United States due to heart disease. That’s readily over ten times the car-related deaths of 40,000 each year in the U.S.
Why not go after the medical and healthcare field and assert that if they all openly shared their proprietary efforts it would save lives?
If you ponder that question, you’ll see that once again the rebuttal arguments mentioned herein about doing so for self-driving cars are equally applicable.
Shifting gears, another aspect to consider is that the efforts toward conventional cars becoming safer will presumably continue to reduce the annual number of deaths and injuries of car accidents.
Step by step, by adding additional safety-related features to conventional cars, the impact of car crashes can be lessened.
This has the potential of closing the gap toward the advantages that true self-driving cars might provide.
In essence, suppose that conventional cars become “safer” and that the number of deaths and injuries is reduced by some percentage Y. If the basis for wanting to open-up self-driving car tech is that it is intended to save lives, the conventional approach might have already lessened to some degree the volume involved (note that all lives are precious and so it is awkward and difficult to argue about net lives savings, though it is a harsh reality when considering these kinds of matters).
For more details about ODDs, see my indication at this link here: https://www.aitrends.com/ai-insider/amalgamating-of-operational-design-domains-odds-for-ai-self-driving-cars/
On the topic of off-road self-driving cars, here’s my details elicitation: https://www.aitrends.com/ai-insider/off-roading-as-a-challenging-use-case-for-ai-autonomous-cars/
I’ve urged that there must be a Chief Safety Officer at self-driving car makers, here’s the scoop: https://www.aitrends.com/ai-insider/chief-safety-officers-needed-in-ai-the-case-of-ai-self-driving-cars/
Expect that lawsuits are going to gradually become a significant part of the self-driving car industry, see my explanatory details here: https://aitrends.com/selfdrivingcars/self-driving-car-lawsuits-bonanza-ahead/
Don’t misinterpret this analysis to somehow suggest that automakers ought to work only in isolation of each other.
That’s not the point.
In fact, as earlier mentioned, there is a continual effort going on of various automakers forming into teams with other automakers, though as can be seen by the tensions often involved that it is a lot harder to do than it might seem on paper.
Plus, there is a lot of hiring of AI self-driving car developers that go across the street from one automaker to the other, thus in a more subtle way perhaps bringing about sharing (albeit one that often generates lawsuits involving claims of having “stolen” proprietary matters).
There are lots of conferences and events about self-driving cars that bring together disparate automakers, allowing for a narrow kind of sharing. In addition, there is the academic world and its efforts on self-driving car advances that frequently are relatively openly being shared.
And, for those of you that have kept up with the movement toward open-source, there have been some efforts in the self-driving car realm toward leveraging open-source approaches (examples include openpilot, apollo, carla, etc.). This though is generally of a limited nature and scope, and not on par with the starry-eyed overarching notion being floated.
All told the wishful thinking that all automakers and self-driving car tech firms ought to make available whatever they have is a bit fantastical and utterly unrealistic.
For those that harbor this impossible dream, I wish you well but gently urge that your energies might be better spent in other ways.
Of course, if we could create AI that’s super-intelligent, maybe we would ask it to devise AI self-driving cars for us. Thus, rather us humans right now myopically focusing on achieving AI-based self-driving cars, let’s instead aim at developing super-intelligent AI that can get the job done for us.
Solved that problem.
Copyright 2020 Dr. Lance Eliot
This content is originally posted on AI Trends.
[Ed. Note: For reader’s interested in Dr. Eliot’s ongoing business analyses about the advent of self-driving cars, see his online Forbes column: https://forbes.com/sites/lanceeliot/]