By Lance Eliot, the AI Trends Insider
There is a famous allegory called the Upstream Parable that provides numerous valuable lessons and can be gainfully applied to the advent of AI autonomous self-driving cars.
The Upstream Parable sometimes referred to as the Rivers Story, has been attributed to various originating sources, including that some suggest it was initially brought up in the 1930s by Saul Alinksy, political activist, and then later by Irving Zola, medical sociologist, though it was perhaps given its greatest impetus via a paper by John McKinlay in 1975 that applied the parable to the domain of healthcare.
I’ll start with a slimmed-down version of the story.
You are walking along the bank of a rushing river when you spy a person in the water that seems to be drowning. Heroically, you leap into the water and save the person. A few minutes later, another person floats by that seems to be drowning. Once again, you jump into the river and save the person.
This keeps happening, again and again.
In each case, you dive in, and though you manage to save the person each such time, doing so denies you the chance to go upstream and ascertain why all these people are getting into the water to begin with, for which you might be able to bring the matter to an overall halt and prevent anyone else from further getting into the dangerous waters.
And that’s the end of the story.
You might be thinking, what gives with this?
Why is it such a catchy parable?
By most interpretations, the story offers a metaphor about how we oftentimes are so busy trying to fix things that we don’t pay attention to how they were originating. Our efforts and focus go toward that which we immediately see. And, especially when something is demanding incessantly our rapt attention right away. If you can take a breather and mull things over in such a situation, you might ultimately be able to solve the matter entirely by going upstream, make a fix there, rather than being battered over and over downstream.
In fact, it could be that one fix at the upstream would prevent all the rest of the downstream efforts, meaning that economically it is potentially a lot more sound to deal with the upstream rather than the frenetic and costly downstream activities.
This can be applied to healthcare in a myriad of ways. For example, suppose that a populace has improper hygiene habits and lives in a manner that encourages disease to take hold. Upon arriving at such a locale, your first thought might be to build a hospital to care for the sick. After a while, the hospital may fill up, so you need to build another hospital. On and on, this merry-go-round goes, devoting more and more resources to building hospitals to aid the ill.
It would be easy to fall into the mental trap of putting all your attention toward those hospitals.
You might chew-up your energy on dealing with:
- Are the hospitals running efficiently?
- Do hospitals have sufficient medical equipment?
- Can you keep enough nurses and doctors on-staff to handle the workloads?
Recalling the lesson of the Upstream Parable, maybe there ought to be attention given to how the populace is living and try to find ways to cut down on the breaking out of disease. That’s upstream and it is the point at which the production of ill people is taking place. Imagine, if you did change the upstream to clean things up and prevent or at least reduce by a large measure the rampant disease, you’d no longer need such a large volume of hospitals, and nor all that equipment, and nor have the issues of staffing the medical teams in a large-scale way.
Notice too that everyone involved in the matter is doing what they believe best to do.
In other words, those building all those hospitals perceive a need to heal the sick, and so they are sincerely and genuinely “doing the right thing.” Unfortunately, they are consumed mightily by that task, akin to pulling drowning people out of the rushing river, and thus they fail to consider what’s upstream and potentially better ways to “cure” the people of their ills.
Okay, that’s the overarching gist of the upstream and downstream related fable.
There are numerous variants of how the story is told.
Some like to say that the persons falling into the water are children and that you are therefore saving essentially helpless children (and, as though to go even further, sometimes the indication is that they are babies).
I guess that might make the parable more engaging, but it doesn’t especially change the overall tenor of the lessons involved.
Here’s one reason that some like to use children or babies in place of referring to adults.
A bizarrely distorted reaction by some is that if it is adults that are falling into the water, why aren’t they astute enough to stop doing so, and why should it be that anyone else should be worried about saving adults that presumably should know better (thus, substituting children or babies makes that less arguable, but I must say that the somewhat cynical and bitter portrayal of adults is a bit alarming since it could be that something beyond their power is tossing them into the drink, and anyway it fights against the spirit of the parable overall).
Another variation of the story has a second individual that comes to aid in saving the drowning subjects.
At the end of the story, this second individual, after having helped to pull person after person out of the river, suddenly stops doing so and walks upstream.
The first individual, still steeped in pulling people out of the water, yells frantically to the second individual, imploring with grave concern, where are they going?
I’m going upstream to find out what’s going on and aim to stop whoever is tossing people into the river, says the second individual.
End of story.
That’s a nifty variant.
Well, in the first version, the person saving the lives has no chance to do anything but continue to save lives (we can reasonably conclude that if the saving were to be curtailed, person after person would drown).
In the second version, we hope or assume that the first individual can sufficiently continue to save lives, while the second person scoots upstream to try and do something about the predicament.
Of course, life is never that clear cut.
It could be that the second person leaving will lamentably present a serious and life-denying result at the downstream saving-lives position.
In which case, we need to ponder as to whether it is better to keep saving lives in the immediate, rather than trying to solve the problem overall, or that you must make a death sentence decision to essentially abandon some to their deaths to deal with the problem by sorting out its root.
On a related topic, nearly all seasoned software developers and AI builders tend to know that whenever you have a budding system that is exhibiting problems, you seek to find the so-called root cause.
If you spend all your time trying to fix errors being generated by the root cause, you’ll perpetually be in a bind of just fixing those errors and never stop the flow.
Anyway, the variant to the parable is quite handy since it brings up a devilish dilemma.
While in the midst of dealing with a crisis, can you spare time and effort toward the root cause, or would that meanwhile generate such adverse consequences that you are risking greater injury by not coping with the immediate and direct issues at-hand?
Keep in mind too that just because the second person opts to walk upstream, we have no way of knowing whether the upstream exploration will even be successful.
It could be that the upstream problem is so distant that the second individual never gets there, and in which case, if meanwhile, people were drowning, it was quite a hefty price to pay for having not solved the root problem.
Or, maybe the second individual finds the root, but they are unable to fix it quickly (maybe it’s a troll that is too large to battle, and instead the second individual has to try and prevent people from wandering into its trap, but this only cuts down on say one-third of the pace of people getting tossed into the river).
This means that for some time, those drowning are going to keep drowning.
Here’s an even sadder possibility.
The second individual reaches the upstream root and tries to fix the problem, yet somehow, regrettably, makes it worse (maybe it was a bridge that people were falling off, and while attempting to fix the bridge, the second individual messed-up and the bridge is even more precarious than it was before!).
It could be that up until then, the first individual was able to keep up with saving those drowning, and now, ironically, after the second individual tried to fix the problem, and in the meantime wasn’t around to help save the drowning victims, there are a slew more people falling into the water, completely overwhelming the first individual.
As you can see, I like this latter version that includes the second individual, allowing us to extend the lessons that can be readily gleaned from the parable.
Some though prefer using the simpler version.
It all depends upon the point that you are trying to drive home by using the tale.
For those of you that are smarmy, I’m sure that you’ve already come up with other variations.
Why not make a net that is stretched across the river and catches all those people?
There, problem solved, you proudly proclaim.
Well, which problem?
The problem of the people drowning at the downstream position, or the problem of the people being tossed into the river and possibly leading to being drowned (hopefully, they don’t drown before they reach your net).
In any case, yes, it might be sensible to come up with a more effective or efficient way to save the drowning persons.
That doesn’t necessarily negate the premise that it is the root that deserves attention, but I appreciate that you’ve tried to find a means to reduce the effort at the downstream, which maybe frees up those that are aiming to go upstream to find and fix the root cause.
One other last facet to mention, and it somewhat dovetails into the notion of creating and putting in place the net, sometimes there is such a massive setup of infrastructure at the downstream that it becomes unwieldy and takes on a life of its own to deal with.
Furthermore, and the twist upon a twist, suppose that the net gets nearly all, but a few happen to go underwater and aren’t saved by the net.
Imagine someone standing downstream of the (already) downstream net.
They might end up in the same parable, and upon coming up to find you and your net, believe they have found the root cause.
It could be that the root cause is further upstream and that there are lots of other intervening downstream solutions, all of which are (hopefully) mitigating the upstream, yet it might be difficult to figure out what’s the root versus what’s not the root.
There could be a nearly infinite series of downstream solutions, all well-meaning, each of which makes the whole affair incredibly complex and confounding, while there might be an elegant end to the monstrosity by somehow getting to the real root.
Well, that was quite an instructive look at the fable.
You might be wondering, can the fable be used in other contexts, such as something AI-related (that’s why I’m here).
Yes, indeed, here’s an interesting question to ponder: “Will the advent of AI-based true self-driving cars potentially find itself getting mired in downstream matters akin to the Upstream Parable?”
Let’s unpack the matter and see.
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 AI-based 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, which 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, the public must 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 The Parable
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.
Sounds pretty good.
No need for any arcane fables or tall tales.
But, wait, give the Upstream Parable a chance.
Some today are arguing that more regulation is needed at the federal level to guide how self-driving cars will be designed, built, and fielded.
Those proponents tend to say that having the states or local authorities in cities and counties having to come up with guidelines for the use of self-driving cars is counterproductive.
You might be surprised to know that many of the automakers and self-driving tech firms seem to generally agree with the notion that the guidelines ought to be at the federal level.
One reason would be the presumed simplicity of having an across-the-board set of rules, rather than having to adjust or craft the AI system and driverless car to accommodate a potential morass of thousands upon thousands of varying rules across the entire country.
On the other hand, a cogent argument is made that having a singular federal level approach might not allow for sufficient flexibility and tailoring that befits the needs of local municipalities.
Let’s suppose that the local approach prevails (I’m not making such a proclamation, it’s just a what-if).
If self-driving cars have trouble coping at the local levels, we might become focused on the downstream matters.
Meanwhile, one might contend that it was the upstream that needed to provide an overarching approach that was sufficient to abate the downstream issues.
Back to the parable we go.
Suppose a fleet of self-driving cars is owned by a particular automaker.
The self-driving cars communicate with a cloud-based system, via OTA (Over-The-Air) electronic capabilities, and pull down patches and updates to the AI system that’s on-board, and also the on-board system uploads collected sensory data and other info from the self-driving car.
Pretend that something goes awry in the self-driving cars of that fleet.
Do you try to quickly deal with each individual self-driving car, which might be on the roadway and endangering passengers, pedestrians, or other human-driven cars, or do you try to ferret out the root cause and then see if you can get that patch shoved out to the fleet in-time?
Some assert that this very kind of issue is why there ought to be a kill button or kill switch inside all self-driving cars, allowing presumably for a human passenger to make a decision right there in the driverless car to stop it from processing.
In any case, you could liken this to the upstream versus downstream fable.
Pleasingly, once again, lessons are revealed due to a handy underlying schema or template.
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: http://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/
Generally, the Upstream Parable is pretty handy for lots of circumstances.
Part of the reason it is so memorable is due to the aspect that it captures innately what we see every day, and helps to bring to light the otherwise hidden or unrealized elements of systems around us that we are immersed in.
While standing at the DMV and waiting endlessly to get your driver’s license renewed, you have to let your mind wander to keep your sanity and wonder whether you’ve found yourself floating in the downstream waters.
Drowning in paperwork!
If the DMV had its act together, there’d be a solution at the root that would make your desire to renew your driver’s license a bit less arduous and frustrating.
For sanity sake, go ahead and use the fable to your heart’s content and keep finding ways to balance the downstream with the upstream, aiming to prevent problems before they arise and make the world a better place.
That’s a good lesson no matter how you cut it.
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/]