Let’s explore the myriad of ways that the words “wash” and “washing” can be stretched and utilized for communicating a variety of pearls of wisdom.
We know for example that people sometimes forewarn that you should not wash your dirty laundry in public. I remember as a child that adults often cautioned that an improperly stated utterance could lead to getting your mouth washed out with soap. Another oft-cited phrase was that everything seems to ultimately come out in the wash.
If you were worried about something that you didn’t wish to be associated with, the recommended idea was to see if you could wash your hands of it. All sorts of washing-related considerations are commonly bandied around, including that you could be summarily washed up or washed over. Washing the egg off your face is an old adage that seems to still occasionally come up in conversations.
The use of colors to represent variations of washing is also relatively well-known. It is said that the notion of whitewashing traces back to at least the 1500s. There are concerns expressed about red washing, purple washing, and so on. I would dare say that perhaps greenwashing is one of the more frequently used catchphrases these days, referring ostensibly to an act of hollowness when touting sustainability and yet not backing up the asserted talk with any substance of walk-the-talk backbone.
You might not know about one of the newest versions of washing, namely AI Ethics washing.
Some prefer to shorten the phrasing to Ethics Washing, though this can produce some confusion since this seeming alternative wording could refer to just about any kind of ethics-oriented washing. The particular form of Ethics Washing that I’ll be discussing herein consists of AI-focused ethics and all related ethical considerations. For sake of clarity, I’d like to suggest that Ethics Washing covers a wide variety of ethics washing that might have little or nothing to do with AI per se. And that AI Ethics washing is a particular kind of Ethics Washing that aims specifically at the realm of AI.
You might be wondering, what exactly does AI Ethics washing consist of?
My overall definition is that AI Ethics washing can be defined as follows:
- AI Ethics washing entails giving lip service or window dressing to claimed caring concerns about AI Ethics precepts, including at times not only failing to especially abide by Ethical AI approaches but even going so far as to subvert or undercut AI Ethics approaches.
For my ongoing and extensive coverage of AI Ethics and Ethical AI, see the link here and the link here, just to name a few.
A quick example of AI Ethics washing might be illustrative for you.
Suppose a firm that is crafting an AI system is desirous of letting the world know about how great their AI is going to be. The firm decides that one means of garnering a lot of positive press and social media attention about the AI would be to publicize that it is devised to be entirely fair and balanced in how the AI functions. The AI is entirely trustworthy. The firm has strictly adhered to the tenets of creating so-called Responsible AI, see my coverage at the link here. The assertion is that all of the prevailing AI Ethics principles were integrally interwoven into the AI system.
Just one little problem.
Turns out that the firm did none of those things.
They did not abide by AI Ethics precepts. They said that they did, but they did not do so. The leaders of the firm and the marketing team decided that claiming they had stridently observed Ethical AI considerations would be presumably good for business. No need to actually do the hard work of dealing with those pesky AI Ethics guidelines, and instead just say that you did.
Voila, they instantly are able to promote their AI by jumping onto the AI Ethics bandwagon.
But this is a risky path and one that will in fact potentially generate big problems.
Firms and leaders that decide to falsely invoke AI Ethics when they have done little to abide by Ethical AI are potentially setting themselves up for a lot of fallout. First, if they are someday exposed regarding their AI Ethics falsehood they are risking a severe reputational backlash. They lied about being AI Ethics minded. In addition, having been caught in a lie, irrespective of having to do with AI Ethics also gets them into added hot water. It is a twofer of lying.
Second, numerous legal ramifications can bite them and their firm. One is that they didn’t do what they said they did and can be potentially legally held liable for their false claims. Another is that their AI is probably going to end up violating laws involving societally sensitive areas such as exhibiting undue biases and acting in discriminatory ways. The list of legal issues is lengthy and can end up forcing the firm into costly legal battles and possibly sinking the entire ship, as it were.
Why in the heck would a firm and its leaders choose to use AI Ethics washing?
Well, it can be somewhat costly to incorporate Ethical AI practices, though the counterargument is that the cost, in the end, will be readily exceeded by the benefits of having AI that is genuine and of a heightened caliber in adhering to AI Ethics approaches. Nonetheless, some firms would prefer to get their AI out the door soonest, and later on, figure that they will worry about fallout from not having considered Ethical AI during the development process.
The old line seems to come to bear, consisting of pay me now or pay me later. Some leaders and firms figure it is worth rolling the dice and hoping that they won’t have to incur the pay-me-later price when they opt to avert the pay-me-now facets. I would argue that there is no free lunch when it comes to AI Ethics. You either do your part, or you bear the consequences.
That’s not to say that there isn’t a lot of wiggle room in all of this.
Firms might dip their toes into AI Ethics and then try to exaggerate how much they have done. Their potential assumption is that they will have a sufficient defense to counter any accusations that they weren’t incorporating AI Ethics at all. They can point to some form of half-hearted AI Ethics activities that might get them off the hook. Thus, the debate then shifts from their not having done any AI Ethics efforts and instead becomes whether or not they did enough.
This is an argument that can go nearly endlessly and allow an AI Ethics washing purveyor a lot of room to maneuver.
Part of the loosey-goosey aspects is that there aren’t as yet agreed-to universal and definitively implementable standards about AI Ethics. Without a cohesive and comprehensive set of metrics, any discussions about whether AI Ethics was being appropriately observed will be tenuous and muddied. The firm will insist they did enough. An outsider or other claiming that the firm didn’t do enough will have an uphill battle showcasing such a counter contention. Ambiguity can reign.
Before getting into some more meat and potatoes about the wild and woolly considerations underlying the AI Ethics washing, let’s establish some additional fundamentals on profoundly integral topics. We need to briefly take a breezy dive into AI Ethics and especially the advent of Machine Learning (ML) and Deep Learning (DL).
You might be vaguely aware that one of the loudest voices these days in the AI field and even outside the field of AI consists of clamoring for a greater semblance of Ethical AI. Let’s take a look at what it means to refer to AI Ethics and Ethical AI. On top of that, we will explore what I mean when I speak of Machine Learning and Deep Learning.
One particular segment or portion of AI Ethics that has been getting a lot of media attention consists of AI that exhibits untoward biases and inequities. You might be aware that when the latest era of AI got underway there was a huge burst of enthusiasm for what some now call AI For Good. Unfortunately, on the heels of that gushing excitement, we began to witness AI For Bad. For example, various AI-based facial recognition systems have been revealed as containing racial biases and gender biases, which I’ve discussed at the link here.
Efforts to fight back against AI For Bad are actively underway. Besides vociferous legal pursuits of reining in the wrongdoing, there is also a substantive push toward embracing AI Ethics to righten the AI vileness. The notion is that we ought to adopt and endorse key Ethical AI principles for the development and fielding of AI doing so to undercut the AI For Bad and simultaneously heralding and promoting the preferable AI For Good.
On a related notion, I am an advocate of trying to use AI as part of the solution to AI woes, fighting fire with fire in that manner of thinking. We might for example embed Ethical AI components into an AI system that will monitor how the rest of the AI is doing things and thus potentially catch in real-time any discriminatory efforts, see my discussion at the link here. We could also have a separate AI system that acts as a type of AI Ethics monitor. The AI system serves as an overseer to track and detect when another AI is going into the unethical abyss (see my analysis of such capabilities at the link here).
In a moment, I’ll share with you some overarching principles underlying AI Ethics. There are lots of these kinds of lists floating around here and there. You could say that there isn’t as yet a singular list of universal appeal and concurrence. That’s the unfortunate news. The good news is that at least there are readily available AI Ethics lists and they tend to be quite similar. All told, this suggests that by a form of reasoned convergence of sorts that we are finding our way toward a general commonality of what AI Ethics consists of.
First, let’s cover briefly some of the overall Ethical AI precepts to illustrate what ought to be a vital consideration for anyone crafting, fielding, or using AI.
For example, as stated by the Vatican in the Rome Call For AI Ethics and as I’ve covered in-depth at the link here, these are their identified six primary AI ethics principles:
- Transparency: In principle, AI systems must be explainable
- Inclusion: The needs of all human beings must be taken into consideration so that everyone can benefit, and all individuals can be offered the best possible conditions to express themselves and develop
- Responsibility: Those who design and deploy the use of AI must proceed with responsibility and transparency
- Impartiality: Do not create or act according to bias, thus safeguarding fairness and human dignity
- Reliability: AI systems must be able to work reliably
- Security and privacy: AI systems must work securely and respect the privacy of users.
As stated by the U.S. Department of Defense (DoD) in their Ethical Principles For The Use Of Artificial Intelligence and as I’ve covered in-depth at the link here, these are their six primary AI ethics principles:
- Responsible: DoD personnel will exercise appropriate levels of judgment and care while remaining responsible for the development, deployment, and use of AI capabilities.
- Equitable: The Department will take deliberate steps to minimize unintended bias in AI capabilities.
- Traceable: The Department’s AI capabilities will be developed and deployed such that relevant personnel possesses an appropriate understanding of the technology, development processes, and operational methods applicable to AI capabilities, including transparent and auditable methodologies, data sources, and design procedure and documentation.
- Reliable: The Department’s AI capabilities will have explicit, well-defined uses, and the safety, security, and effectiveness of such capabilities will be subject to testing and assurance within those defined uses across their entire lifecycles.
- Governable: The Department will design and engineer AI capabilities to fulfill their intended functions while possessing the ability to detect and avoid unintended consequences, and the ability to disengage or deactivate deployed systems that demonstrate unintended behavior.
I’ve also discussed various collective analyses of AI ethics principles, including having covered a set devised by researchers that examined and condensed the essence of numerous national and international AI ethics tenets in a paper entitled “The Global Landscape Of AI Ethics Guidelines” (published in Nature), and that my coverage explores at the link here, which led to this keystone list:
- Justice & Fairness
- Freedom & Autonomy
As you might directly guess, trying to pin down the specifics underlying these principles can be extremely hard to do. Even more so, the effort to turn those broad principles into something entirely tangible and detailed enough to be used when crafting AI systems is also a tough nut to crack. It is easy to overall do some handwaving about what AI Ethics precepts are and how they should be generally observed, while it is a much more complicated situation in the AI coding having to be the veritable rubber that meets the road.
The AI Ethics principles are to be utilized by AI developers, along with those that manage AI development efforts, and even those that ultimately field and perform upkeep on AI systems. All stakeholders throughout the entire AI life cycle of development and usage are considered within the scope of abiding by the being-established norms of Ethical AI. This is an important highlight since the usual assumption is that “only coders” or those that program the AI is subject to adhering to the AI Ethics notions. As earlier stated, it takes a village to devise and field AI, and for which the entire village has to be versed in and abide by AI Ethics precepts.
Let’s also make sure we are on the same page about the nature of today’s AI.
There isn’t any AI today that is sentient. We don’t have this. We don’t know if sentient AI will be possible. Nobody can aptly predict whether we will attain sentient AI, nor whether sentient AI will somehow miraculously spontaneously arise in a form of computational cognitive supernova (usually referred to as the singularity, see my coverage at the link here).
The type of AI that I am focusing on consists of the non-sentient AI that we have today. If we wanted to wildly speculate about sentient AI, this discussion could go in a radically different direction. A sentient AI would supposedly be of human quality. You would need to consider that the sentient AI is the cognitive equivalent of a human. More so, since some speculate we might have super-intelligent AI, it is conceivable that such AI could end up being smarter than humans (for my exploration of super-intelligent AI as a possibility, see the coverage here).
Let’s keep things more down to earth and consider today’s computational non-sentient AI.
Realize that today’s AI is not able to “think” in any fashion on par with human thinking. When you interact with Alexa or Siri, the conversational capacities might seem akin to human capacities, but the reality is that it is computational and lacks human cognition. The latest era of AI has made extensive use of Machine Learning (ML) and Deep Learning (DL), which leverage computational pattern matching. This has led to AI systems that have the appearance of human-like proclivities. Meanwhile, there isn’t any AI today that has a semblance of common sense and nor has any of the cognitive wonderment of robust human thinking.
ML/DL is a form of computational pattern matching. The usual approach is that you assemble data about a decision-making task. You feed the data into the ML/DL computer models. Those models seek to find mathematical patterns. After finding such patterns, if so found, the AI system then will use those patterns when encountering new data. Upon the presentation of new data, the patterns based on the “old” or historical data are applied to render a current decision.
I think you can guess where this is heading. If humans that have been making the patterned upon decisions have been incorporating untoward biases, the odds are that the data reflects this in subtle but significant ways. Machine Learning or Deep Learning computational pattern matching will simply try to mathematically mimic the data accordingly. There is no semblance of common sense or other sentient aspects of AI-crafted modeling per se.
Furthermore, the AI developers might not realize what is going on either. The arcane mathematics in the ML/DL might make it difficult to ferret out the now hidden biases. You would rightfully hope and expect that the AI developers would test for the potentially buried biases, though this is trickier than it might seem. A solid chance exists that even with relatively extensive testing that there will be biases still embedded within the pattern matching models of the ML/DL.
You could somewhat use the famous or infamous adage of garbage-in garbage-out. The thing is, this is more akin to biases-in that insidiously get infused as biases submerged within the AI. The algorithm decision-making (ADM) of AI axiomatically becomes laden with inequities.
Let’s now return to the topic of AI Ethics washing.
There are four major variants of AI Ethics washing that I usually see occurring (I’ll explain these in a moment):
1) The AI Ethics Washers That Don’t Know They Are: AI Ethics washing by ignorance or illiteracy about AI and/or AI Ethics
2) The AI Ethics Washers That Fall Into It: AI Ethics washing by inadvertent slippage though otherwise genuine about AI Ethics and AI
3) The AI Ethics Washers That Thinly Stretch: AI Ethics washing by purposeful intent though just by a smidgeon and at times nearly excusable (or not)
4) The AI Ethics Washers That Know And Brazenly Peddle It: AI Ethics washing all-out and by insidious and often outrageous design
I would generally suggest that the four variants range from the shall we say the most innocent to the guiltiest, as to the awareness of what AI Ethics washing is. Let’s walk through each of the four, starting with the first one and making our way to the rather shameful fourth one.
First, you’ve got those that are somewhat of the unwashed in that they do not know what AI Ethics is, they don’t know what AI Ethics washing is, and they probably don’t even know much about AI either. You could say they are ignorant or illiterate on those topics. For them, they are probably committing AI Ethics washing and blindly and blissfully do not realize that they are doing so.
This is sad.
It can be especially bad too if the AI Ethics washing is being done by a major news agency or high-profile social media influencer. They might have been fed a bed of lies, and they did not vet those falsehoods. Meanwhile, they use their reach and influence to perpetuate the AI Ethics washing claims. Sad and worse still that it does a disservice to society all told. Shame on those that let themselves be fooled. They need to wise up. Remember, being fooled and looking foolish are close cousins.
Next in order is the AI Ethics washing which is a slipup. Imagine that a firm has been doing quite well in abiding by AI Ethics precepts. We can congratulate them for doing this. Unfortunately, suppose at some juncture they make an announcement about their AI that is not well-supported from an AI Ethics viewpoint. If this is a relatively innocuous statement or unintentional error, we might grant them some latitude. Of course, if the point they made is egregiously over the line, the slippage is not so easily overlooked. There is a famous line that it takes forever to build a reputation and yet only takes a brief moment to utterly demolish it.
We next enter into the last two of the four categories.
These are the culprits that are fully aware of AI Ethics washing and consciously with overt intent decide to utilize it, perhaps as part of a corporate strategy or by other divined means. The main difference between these latter two is that the AI Ethics washing might be of a minor nature, or it might be of a significant and crucial nature. You’ve got some that opt to stretch things and just slightly edge over the line. There are others that are whole hog willing to take AI Ethics washing to the extreme.
You might be thinking that certainly any of the extreme AI Ethics washing would have to be obvious and that the extremist would get caught with their hand in the cookie jar. Everyone would see that the emperor has no clothes. Lamentedly, given the overall confusion about AI and AI Ethics in the world today, there is enough murkiness that even the extreme AI Ethics washing can get a free pass.
This can be quite irksome to those that side with being serious and sober about AI Ethics. They watch as someone else tosses around all manner of slop that is exceedingly evident in AI Ethics washing. The extremist gets massive media attention. They have their proverbial 15 minutes of fame. Those that are doing the real work and the right thing when it comes to AI Ethics can feel rightfully exasperated and justifiably upset when AI Ethics washing is undertaken scot-free by others in the marketplace.
It can be almost likened to juicing and doping that takes place in sports. An athlete that has put their heart and soul into naturally becoming a top-notch athlete can be entirely crestfallen if someone else manages to compete at their same level and does so via the use of banned performance-enhancing drugs. Should you call out the other spoiler? Should you perhaps quietly opt to also take those drugs, fighting fire with fire? This is a conundrum. For my discussion about how juicing or doping happens in the AI field, see the link here.
Now that we’ve covered a bit about AI Ethics washing, we can introduce a bunch of other related catchphrases that are equally in the same arena.
Here are a few that we can briefly explore:
- AI Ethics theatre
- AI Ethics shopping
- AI Ethics bashing
- AI Ethics shielding
- AI Ethics fairwashing
Let’s briefly examine each of those catchphrases. Not everyone agrees with what each phrase precisely denotes, so I’ll share with you my general impressions.
AI Ethics Theatre
AI Ethics theatre can be somewhat similar to AI Ethics washing in that the idea is to make a sizable showing of having abided by AI Ethics precepts in a considered staged and ceremonial way. If the firm that is carrying out the circus or AI Ethics theatre really was abiding by AI Ethics practices, you can make the argument that they justly ought to be able to do so. Indeed, you could further contend that this will hopefully inspire others to also abide by AI Ethics.
On the other hand, it would seem that AI Ethics theatre tends to usually go overboard. The circus act of having all those ponies and elephants can tend to overstate what was truly undertaken. This in turn begins to enter into the second, third, or fourth category of the aforementioned AI Ethics washing takes. Whether the theatre is more good than bad (such as being inspirational), or more bad than good (possibly spurring AI Ethics washing by others), remains to be seen.
AI Ethics Shopping
Envision that a company is going to build an AI system and realizes that they should include AI Ethics aspects during the development life cycle of the AI. Which of the many AI Ethics guidelines should they make use of?
Picking several of them at once could be confusing and make their AI efforts overly bulky. The odds are that the development effort will be more likely to abide by AI Ethics practices if there is a single internally adopted set that all can readily refer to and make sense of.
Okay, so one means to land on an AI Ethics set of principles would be to pick just one of the many available. Another would be to take several sets and try to merge them together. The problem with the merging could be that you expend a lot of precious energy and time on debating about how to best merge the sets into one comprehensive whole. That kind of attention would probably be diverting you from getting on with the development process, plus might get the AI team riled up merely over the acrid debates that might have occurred during the AI Ethics merging activity.
Overall, you might also try to steer your attention to an AI Ethics set of guidelines that you believe will be the easiest to adopt. This would seem perfectly fine. Why make life harder than it probably already is? In that same breath, supposing that you pick an AI Ethics set that is watered down. You are aiming to do the least that you can do. You want to hold your head high that you abided by AI Ethics precepts, meanwhile secretly choosing the minimalist or maybe even less so of the bunch.
Some would refer to this as AI Ethics shopping.
You are shopping around to find the AI Ethics principles that will give you the easiest path to claiming that you are abiding by AI Ethics. This does make sense in that why should you have to do more than is necessary? This though can be warped by finding a thinly veiled set of AI Ethics and clinging to it as though it is robust and bona fide when the reality is that it is sparse and marginal.
AI Ethics Bashing
The notion of AI Ethics bashing is somewhat straightforward.
You bash or denigrate the nature of and use of AI Ethics precepts. A common form of bashing would be to insist that AI Ethics guidelines are worthless and not worth the paper they are printed on. Another popular bash is that AI Ethics is an academic exercise that has nothing to do with the real world. There is even the exhortation that AI Ethics is bad, presumably because it provides a false cover for those that want to seem as though they are developing AI properly. In that manner, AI Ethics is a cover-up scheme.
I won’t belabor the AI Ethics bashing phenomena and suggest you see my coverage of why those bashes are either wrong or at least misguided, see the link here.
AI Ethics Shielding
AI Ethics shielding usually refers to the idea that AI Ethics is a deceptive kind of shield that can hide or obscure bad actors and bad AI Ethics efforts. I’ve alluded to this repeatedly throughout this herein discussion.
There is an ongoing qualm that some will proudly display that they are using AI Ethics and the underlying reality is that they are doing almost nothing of the kind. For those that say we should get rid of the shield entirely, I tend to retort that this is akin to throwing out the baby with the bathwater.
AI Ethics Fairwashing
Fairness is important.
You might recall that when I was discussing the AI Ethics precepts, one of the most frequently identified AI Ethics guidelines consists of trying to ensure that the AI is fair or exhibits a semblance of fairness. This has given rise to a catchphrase of “fairwashing” which is used at times to evoke the possibility that an AI system is said or claimed to be fair when it might not be fair or there is little evidence to support that it is fair. This is a bit of a mashup of AI Ethics washing with the conceptual consideration of AI being fair, ergo the shortcut way to express this is by saying that there is potentially fairwashing that can happen. Researchers describe the matter this way: “In particular, due to the growing importance of the concepts of fairness in machine learning, a company might be tempted to perform fairwashing, which we define as promoting the false perception that the learning models used by the company are fair while it might not be so” (by Ulrich Aiıvodji, Hiromi Arai, Olivier Fortineau, Sebastien Gambs, Satoshi Hara, and Alain Tapp in “Fairwashing: The Risk Of Rationalization”).
There is another twist about AI Ethics washing that you ought to be considering.
On a somewhat macroscopic scale, there is an expressed concern that the rise of AI Ethics is a shield or cover for something even grander. You might be aware that there are numerous efforts underway to establish laws about the governance of AI. This is happening all across the globe. Strident efforts are underway in the EU, which I’ve covered in my columns, and likewise in the U.S., along with taking place in many countries.
Some suggest that the embracing of AI Ethics might be a means of staving off the enactment of those laws. Companies can seemingly persuasively argue that new laws are not needed because the use of AI Ethics suitably deals with any AI issues. AI Ethics are usually classified as being a form of “soft law” and are customarily voluntary (all else being equal). Laws about AI are classified as so-called “hard laws” and are not of a voluntary construct (generally).
It is commonly said that firms would abundantly prefer the soft laws over the hard laws, giving them more latitude and leeway. Not everyone agrees with that sentiment. Some say that the soft laws allow firms to get away with inappropriate efforts and the only way to nail them is by enacting hard laws. Others point out that firms would at times prefer hard laws, which can provide a clearer playing field. Hard laws can potentially make all players abide by the same rules. Soft laws allow a kind of pick and choose, thus creating confusion and disrupting the playing field.
Here’s how research depicts AI Ethics washing amidst this grander view of what might be taking place: “On the one hand, the term has been used by companies as an acceptable façade that justifies deregulation, self-regulation, or market-driven governance, and is increasingly identified with technology companies’ self-interested adoption of appearances of ethical behavior. We call such growing instrumentalization of ethical language by tech companies “ethics washing.” Beyond AI ethics councils, ethics washing includes other attempts at simplifying the value of ethical work, which often form part of a corporate communications strategy: the hiring of in-house moral philosophers who have little power to shape internal company policies; the focus on humane design – e.g. nudging users to reduce time spent on apps – instead of tackling the risks inherent in the existence of the products themselves; the funding of work on ‘fair’ machine learning systems which positively obscures deeper questioning around the broader impacts of those systems on society” (by Elettra Bietti, “From Ethics Washing to Ethics Bashing: A View on Tech Ethics from Within Moral Philosophy,” Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency).
At this juncture of this weighty discussion, I’d bet that you are desirous of some illustrative examples that might showcase this topic. There is a special and assuredly popular set of examples that are close to my heart. You see, in my capacity as an expert on AI including the ethical and legal ramifications, I am frequently asked to identify realistic examples that showcase AI Ethics dilemmas so that the somewhat theoretical nature of the topic can be more readily grasped. One of the most evocative areas that vividly presents this ethical AI quandary is the advent of AI-based true self-driving cars. This will serve as a handy use case or exemplar for ample discussion on the topic.
Here’s then a noteworthy question that is worth contemplating: Does the advent of AI-based true self-driving cars illuminate anything about AI Ethics washing, and if so, what does this showcase?
Allow me a moment to unpack the question.
First, note that there isn’t a human driver involved in a true self-driving car. Keep in mind that true self-driving cars are driven via an AI driving system. There isn’t a need for a human driver at the wheel, nor is there a provision for a human to drive the vehicle. For my extensive and ongoing coverage of Autonomous Vehicles (AVs) and especially self-driving cars, see the link here.
I’d like to further clarify what is meant when I refer to true self-driving cars.
Understanding The Levels Of Self-Driving Cars
As a clarification, 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 Level 4 and Level 5 (see my explanation at this link here), while a car that requires a human driver to co-share the driving effort is usually considered at 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-ons 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, 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 contend, see my coverage at this link here).
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 needs to be forewarned about a disturbing aspect that’s been arising lately, namely that despite 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 AI Ethics Washing
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.
One aspect to immediately discuss entails the fact that the AI involved in today’s AI driving systems is not sentient. In other words, the AI is altogether a collective of computer-based programming and algorithms, and most assuredly not able to reason in the same manner that humans can.
Why is this added emphasis about the AI not being sentient?
Because I want to underscore that when discussing the role of the AI driving system, I am not ascribing human qualities to the AI. Please be aware that there is an ongoing and dangerous tendency these days to anthropomorphize AI. In essence, people are assigning human-like sentience to today’s AI, despite the undeniable and inarguable fact that no such AI exists as yet.
With that clarification, you can envision that the AI driving system won’t natively somehow “know” about the facets of driving. Driving and all that it entails will need to be programmed as part of the hardware and software of the self-driving car.
Let’s dive into the myriad of aspects that come to play on this topic.
First, it is important to realize that not all AI self-driving cars are the same. Each automaker and self-driving tech firm is taking its approach to devising self-driving cars. As such, it is difficult to make sweeping statements about what AI driving systems will do or not do.
Furthermore, whenever stating that an AI driving system doesn’t do some particular thing, this can, later on, be overtaken by developers that in fact program the computer to do that very thing. Step by step, AI driving systems are being gradually improved and extended. An existing limitation today might no longer exist in a future iteration or version of the system.
I hope that provides a sufficient litany of caveats to underlie what I am about to relate.
You have almost certainly seen headlines that proclaim the bold assertion that autonomous vehicles are here and self-driving cars are already perfected. The implication is usually that the autonomy aspects are solved. We have derived AI that is as good as human drivers, possibly even better than humans.
Just to assuredly pop that bubble and set the record straight, this is not yet the case.
We do know that human drivers in the United States get into about 2.5 million car crashes annually, deriving over 40,000 annual fatalities, see my stats at the link here. Anybody of seemingly any reasonable ilk would welcome AI driving systems if they were able to drive as safe or more safely than human drivers. Furthermore, the hope is that we will experience mobility-for-all, allowing those that today are mobility constrained to have AI-driven vehicles that provide ready access for convenient and inexpensive transportation.
Some pundits shockingly go the “extra mile” and make the disgraceful claim that self-driving cars will be uncrashable. This is absolutely nutty and completely false. Worse still, it is setting up heightened expectations that cannot be met. If you can convince the populace that self-driving cars are uncrashable, they will holler and bellow the moment that even one instance of a self-driving car-related crash occurs. For my detailed explanation of why the uncrashable claim is looney and a disservice to society, see my coverage at the link here.
All of these kinds of exaggerations or other falsehoods could be said to be covered by the AI Ethics precepts in that if you are abiding by AI Ethics principles you ought to not be making those types of wild and unsubstantiated claims. Thus, these misrepresentations and untruths are readily within the rubric of AI Ethics washing.
AI Ethics washing associated with autonomous vehicles and self-driving cars is widely and wildly abundant, sadly so. A casual and off-the-cuff search of the Internet will show you zillions of zany and unsupported claims about autonomous vehicles. This isn’t just confined to people that are on their own blogs. Major news agencies get caught up in this. Major companies get caught up in this. Startups get caught up in this. Venture Capital firms get caught up in this. Shareholders get caught up in this. And so on.
I would say with gloomy high confidence that AI Ethics washing in this particular domain is rampant.
A specialized variant of the catchphrase about AI Ethics washing that entails autonomy and autonomous systems is the notion of autonowashing. Here is the author Liza Dixon depicting this: “Adapted for automation, autonowashing is defined as the practice of making unverified or misleading claims which misrepresent the appropriate level of human supervision required by a partially or semi-autonomous product, service, or technology. Autonowashing may also be extended to fully autonomous systems, in cases where system capabilities are exaggerated beyond what can be performed reliably, under all conditions. Autonowashing makes something appear to be more autonomous than it really is. The objective of autonowashing is to differentiate and/or offer a competitive advantage to an entity, through the use of superficial verbiage meant to convey a level of system reliability that is misaligned with the technical capabilities of the system. Autonowashing may also occur inadvertently, when one unknowingly repeats erroneous information about the capabilities of an automated system to another. Autonowashing is a form of disinformation, and it is, in a sense, viral” (Transportation Research Interdisciplinary Perspective, “Autonowashing: The Greenwashing of Vehicle Automation,” 2020).
As a reminder from my earlier indication, there are four major variants of AI Ethics washing that I usually see occurring and they are readily found within the autonomous vehicles field too:
- The AI Ethics Washers That Don’t Know They Are: AI Ethics washing by ignorance or illiteracy about AI and/or AI Ethics in autonomous vehicles
- The AI Ethics Washers That Fall Into It: AI Ethics washing by inadvertent slippage though otherwise genuine about AI Ethics and AI in autonomous vehicles
- The AI Ethics Washers That Thinly Stretch: AI Ethics washing by purposeful intent though just by a smidgeon and at times nearly excusable (or not) in autonomous vehicles
- The AI Ethics Washers That Know And Brazenly Peddle It: AI Ethics washing all-out and by insidious and often outrageous design in autonomous vehicles
In addition, you can readily see instances of the other types of AI Ethics washing and associated washing maladies in the autonomy field, including:
- AI Ethics theatre in autonomous vehicles and self-driving cars
- AI Ethics shopping in autonomous vehicles and self-driving cars
- AI Ethics bashing in autonomous vehicles and self-driving cars
- AI Ethics shielding in autonomous vehicles and self-driving cars
- AI Ethics fairwashing in autonomous vehicles and self-driving cars
AI Ethics washing is all around us. We are bathing in it.
I hope that by bringing your attention to this serious and seemingly neverending matter, you will be able to discern when you are being AI Ethics washed. It might be hard to figure out. Those that do AI Ethics washing can be exceedingly clever and devious.
One handy trick is to mix into the wash a bit of truth that gets intermixed with the falsehoods or exaggerations. Because you might readily detect and agree to the truthful part, you are potentially swayed into believing that the other untruthful or deceptive part is also true. A nifty and nasty form of deception.
Let’s call it hogwash.
Can we wash the mouths out of those that outrightly perform AI Ethics washing?
Sorry to report that it is not as easily done as one might wish for. That being said, just because catching and calling out AI Ethics washing might be arduous and at times be like Sisyphus pushing a heavy boulder up a hill, we need to try.
In case you didn’t know, Zeus had tasked him to do this boulder pushing for eternity and the immense rock would forever roll back down once the effort had arrived at the top. I think we are facing the same predicament when it comes to AI Ethics washing.
There will always be more AI Ethics washing that needs to get washed out. That’s a surefire guarantee with nothing wishy-washy whatsoever about it.