Invasive Curve and AI Self-Driving Cars

1161

By Lance Eliot, the AI Trends Insider

The puff adder, a venomous snake, became an invading species in Florida after it was imported into the state as a pet but then was let loose into the local native environment. The Asian long horned beetle was inadvertently brought into the United States when it hitchhiked on imported agricultural products, and now is destroying forests in the northeastern parts of the U.S. The Giant African Land Snail (GALS) loves to eat plants and was inadvertently brought into Florida where it began to threaten the agricultural well being of the state.

We are under siege!

Plants and animals that are not native to the United States are often introduced into the United States and then are able to go on an invasive torrent. At times the invaders are intentionally brought in, such as the puff adder that some thought would be a great pet, while in other cases the invader sneaks in, such as the Asian long horned beetle that hitched a ride to here via imported agricultural products. Our efforts as humans has allowed for new alien species to find an ecological environment that allows them to flourish.

In some cases, we even help the alien species. We might transport the alien species across the country and allow it to readily get started into new neighborhoods. We become the dispersal agent. Whether it is an animal or plant that we carry in our planes, trains, boats, or via other means, we aid something that otherwise could not go large distances on its own, and we put it into new areas for potential invasion. Our awareness that an alien species is taking hold will at times vary from knowing what we are doing and at other times we are quite unaware of what we are doing.

If we begin to realize that we’ve allowed an alien species to gain new ground, there are some occasions where we might welcome the invasion, and other occasions wherein we want to stop it. Our cultural perceptions will partially determine whether we want the alien to continue or we want it to be curtailed. One moment, we might be promoting the alien invader and urging all to adopt it, while later on we might decide that the invader is harmful and decide that it needs to be eradicated. Depopulation of an alien species can be very difficult to undertake and requires various invasive barriers and trickery to slow down or remove.

Some invaders are able to leverage aspects of their new environment to then achieve high abundances rapidly. There might not be any native predators that help keep the invader volume pruned. There might not be any environmental conditions of their original origins that are equaled in the new region, such as a creature that normally cannot survive harsh cold weather, which then becomes unabated in a year-round warm weather climate in a newly implanted region. In fact, there is the enemy release hypothesis that postulates an invader can profit mightily by the avoidance of its otherwise natural enemies when those enemies are not present in a new environment. Humans have even sometimes opted to try and stop an invader by then bringing in a second invader, one that is a natural enemy of the first invader, in hopes of expunging the first invader (and then dealing with the second invader in some other fashion).

Generally, there are three major steps involved in the invasion process:

–          Humans introduce the invader into a new environment

–          Invader spreads in the new environment

–          Invader integrates into the new environment

Of course, it is not always the case that humans are the ones that introduce the invader. It could be that an invader is brought into the United States by the winds and carried across our lands. Or, the invader might arrive by the sea. Overall, though, today’s invaders are most likely being introduced by us humans.

I’ll also point out that though I am focusing on invaders that come into the United States, it is the same kind of invading dynamics that occur for other countries. In other words, a natural plant or animal in the United States can be carried into another country and be considered an invader in that country. Don’t want you to think that I am being myopic and considering that only the United States as somehow the land of invasions. It can happen anywhere.

There’s also another interesting twist to invasion dynamics, namely the possibility of an invasion meltdown. Imagine that an invasive species known as X has established itself into a new ecosystem called A. The invasive species X is a strong invader and takes hold quite firmly in the ecosystem A. This X begins to overtake native species Y and Z. Without the native species Y and Z, another invader Q that otherwise would not have had a chance at establishing itself in the new ecosystem, suddenly begins to flourish. Essentially, the original invader X has produced a sequence of changes that creates an invasion meltdown, and allows invaders to flow into the new ecosystem.  It could be a cascade that becomes nearly impossible to stop as it reaches a widespread pervasive invasion level.

Guess What These Numbers Represent

On the topic of invaders, I’d like to share with you the below numbers. This shows the year and a numeric count. I won’t say just yet what the count represents, and will keep you in suspense until later in this discussion:

1900:            8,000

1910:        458,000

1920:     8,132,000

1930:   23,035,000

1940:   27,466,000

1950:   40,339,000

1960:   61,671,000

1970:   89,224,000

1980: 121,601,000

1990: 133,700,000

2000: 225,821,000

2010: 250,070,000

I’ll let you think about what those counts might be. It seems obvious that the counts are going up. And, there has been a rapid rise, especially at the start, going from a count of 8,000 to a count of 458,000 in just the first ten years of introduction. Over a one hundred and ten years span, the count went from the lowly starting point of 8,000 to become more recently a massive count of 250,070,000. Seems like a pretty successful “invader” upon quick glance.

So, what does all this talk about invasive species have to do with AI self-driving cars?

At the Cybernetic Self-Driving Car Institute, besides developing AI systems for self-driving cars, we also examine the socioeconomic aspects of the emergence of self-driving cars. As such, and this might seem surprising or even shocking to you, we assert that the adoption of self-driving cars could be considered an invasive dynamic.

Yes, it could be that once self-driving cars are introduced into our environment, we might begin to see a phenomenon similar to what happens when an invasive species is introduced into a new environment.

Right now, we have around 260 million conventional non-self-driving cars in the United States. The estimated number of AI self-driving cars in the United States today is tiny in comparison. In fact, none of the AI self-driving cars are yet at the true Level 5 of self-driving cars (which is the point at which self-driving cars are driven by AI as though it was the same as if a human driver was driving the car).  Currently, there are a smattering of maybe Level 4 self-driving cars, mainly being used experimentally, and otherwise the rest of the self-driving cars are somewhere between a Level 2 and a Level 3.

As such, the invasion of the true AI self-driving car has not actually yet begun. But, we believe a true Level 5 self-driving car is possible and that we will ultimately have such cars on our roads, and thus, one could construe the eventual emergence of the true self-driving car as a kind of invader. Presumably, the self-driving car will begin to edge out the non-self-driving car. Some utopian believers seem to think that overnight we’ll all switch from non-self-driving cars to self-driving cars, but realistically it will take a long time for such a switchover to occur. The economic cost to switch cars from today’s pervasive non-self-driving cars to the emerging self-driving cars will prohibit a magical overnight transformation.

Remember when earlier I had stated that humans are the ones that tend to introduce an invading species, well, that’s certainly going to be the case with the true AI self-driving cars. We, the humans, will be introducing the self-driving invader into our ecosystem. I know this kind of sounds like crazy talk, in that an AI self-driving car is not an invading plant or animal, and so you might be wondering how can we think of an inanimate object as an invading species.

Some Argue the Self-Driving Car is Close to a Living Organism

Some would argue that the AI elements of the true self-driving car make it closer to a living organism than if we were referring to say refrigerators as invaders or toothbrushes as invaders. They would also point out that the self-driving cars will have the potential for life and death over us humans, while a toothbrush or refrigerator does not have that same significance.

I don’t think we even need to get into the argument about whether the AI self-driving car is somehow sentient and whether or not we need to consider it as a kind of living thing. Put aside for the moment that whole debate. We can simply consider that we are going to see an introduction of something new into our existing transportation ecosystem, and that the new thing is a self-driving car. From that perspective, we can still make use of the invader dynamics as a framework for considering how this “new species” will take hold.

Indeed, you can consider that there are two viewpoints on the adoption of the true self-driving cars, either you think it is a good thing and we should definitely do it, or you think it is a bad thing and we should consider it a pest that needs to be stopped.

Right now, it seems like the mass media wants us to think that the true self-driving car is a good thing and so it is a welcomed invader that we want to encourage, but the tide might change as we learn more about the recent Uber self-driving car incident that killed a pedestrian (occurred in Tempe, Arizona), and if we see more such incidents arise. There could be a gradual and inexorable loss-of-faith toward the unflagging race to achieve self-driving cars on our public roadways.

Those that are unabashed fans of self-driving cars claims that a true self-driving car won’t get drunk, it won’t get tired, and otherwise will drive a car much better and more safely than humans do.  That’s certainly the hope, but we don’t yet know that’s going to be the case. I am sure that you might say that well, if the true self-driving cars aren’t as good or better than human drivers then we just won’t allow them to proceed, or at least we’ll limit how far they can proceed and until they are perfected.  If you are saying this, I think you have just fallen into the whole point about thinking of the true self-driving car as a potential invader. As mentioned earlier, sometimes humans first believe that an invader is a good thing, and later on change our minds and decide it isn’t a good thing. We then either want to stop it, or maybe at least limit how much it takes hold.

Let me for a moment return to my earlier listing of years and counts. You might still be trying to determine what those counts represent (I gave you a subtle clue when I said that there are about 260 million conventional cars in the United States today).  As indicated, the counts over a one hundred and ten years span went from the starting point of 8,000 to become more recently a count of 250,070,000. If you guessed that it was the introduction of the automobile in the United States, you’d be right! These are figures reported by various United States governmental bodies that keep track of the number of registered automobiles in the United States.

If you were to plot out the counts, you’d see that it makes a kind of curve. Roughly it makes what is often referred to as an invasion curve. An invasion curve starts small and gradually builds, and usually begins to taper off. The tapering is typically due to some form of saturation or symbiotic balance that the invader reaches in the ecosystem.

There are these stages to an invasion curve:

–          Prevention

–          Eradication

–          Containment

–          Resource Protection

For prevention, the invader is not yet introduced and so if you don’t want the invader to take hold then you put up barriers or take steps to prevent it from emerging into the ecosystem. One could say that the legislation today about self-driving cars is a kind of “preventative” measure, at least with regard to trying to ensure that the self-driving cars are safe for being on the roadways.

Once the invader starts to emerge, the next stage is eradication. Here, the population is still relatively small and it is feasible typically to try and eradicate it before it becomes more pervasive. This could happen with self-driving cars in that if the first set of true self-driving cars begins to falter, such as say hitting other cars or hitting pedestrians, we might see efforts to eradicate the self-driving car. Now, that being said, it seems like the self-driving car is going to be one of those invaders that is the proverbial letting the genie out of the bottle. It seems unlikely that one could eradicate per se the self-driving car, and instead more likely just prevent it from continuing to spread until the point at which it is able to be proven that it has been changed to be safe (or safer).

In the containment stage, the invader has had a rapid increase in distribution and has become quite abundant in the ecosystem. This could happen with the true self-driving cars. We might have self-driving cars become widely popular, and then, out-of-the-blue, suppose there is some fatal flaw that had not previously been detected. It could put momentary brakes on the continued expansion of self-driving cars. There would likely be public efforts to contain the rise of the self-driving car.

Finally, once an invader has taken deep roots in the ecosystem, it becomes mainly a matter of resource protection. This is somewhat akin to how conventional non-self-driving cars are today. We have all sorts of rules and regulations that provide for our protection and prevent cars from becoming unsafe or at least less safe than they are today.

I realize that most pundits are focused on the adoption curve when it comes to self-driving cars. The notion is that we want to adopt self-driving cars and so let’s push it along on the stages of adoption. That’s obvious and easy to consider. We’ve provided a somewhat alternative view that another way to consider self-driving cars is as an invading species. As such, it is helpful to consider the stages of an invasion curve and consider what we might do if the invader has issues along the way of its insertion into our ecosystem. Certainly, this is a somewhat unconventional way to think of this, but nonetheless helpful when considering the socioeconomic impacts of the true Level 5 self-driving cars.

As per the Romans, every new beginning comes from some other beginning’s end.

This content is originally posted on AI Trends.