Quantum computers will impact self-driving cars, here's how.
Quantum supremacy achieved!
Quantum supremacy, achieved?
The news recently was agog with the claim that the so-called and highly sought “quantum supremacy” had been achieved via an effort undertaken by Google researchers.
Not everyone agreed though that the Google effort warranted waving the superlative supremacy flag.
That’s not to say that the use of their 54-qubit Sycamore processor wasn’t notable, and in fact, does provide another handy stride toward achieving viable quantum computing, but whether it was the vaunted moment of true supreme magnificence is something that many would argue is premature and supremely debatable.
Let’s first consider what the fuss about quantum computers consists of and then noodle on the supremacy claim.
In addition, it would be useful to ponder how quantum computers will impact the advent of true self-driving cars.
Quantum Computing Explained
Today, our conventional computers pretty much employ the same hardware architecture that they have since the beginning of the computer field.
Sure, the hardware has gotten faster, smaller, and cheaper in cost, yet nonetheless, the underlying design principles and operational approach is still intact.
Often referred to as classical computing, you can also be in-the-know by calling the hardware a Turing machine (named after Alan Turing, a pioneer in mathematics and computing), or at times mumble to your friends and everyday strangers that the hardware is von Neumann architecture (named after physicist and mathematician John von Neumann).
Now that you’ve gotten that groundwork, it is time to introduce Einstein into this equation.
In physics, there is a field of study known as quantum mechanics. At the atomic and subatomic level of our universe and matter, there are aspects going on that seem to be somewhat mysterious and not readily explained by ordinary physics. Those in the field of quantum mechanics offer indubitably fascinating theories to explain the unusual behaviors occurring at the particle and sub-particle levels.
For example, it appears that particles can be somehow connected with each other when at long distances from each other, and yet there’s no apparent reason why the disparate particles should be able to interconnect with each other. Generally, this is known as quantum entanglement.
Einstein was in the middle of the heyday of quantum mechanics formulation. He is famously known for having said that this surprising kind of entanglement is “spooky action at a distance” and fretted over how such an odd phenomenon could be fully explained.
Thus, you can quote Einstein that this whole thing is spooky.
Anyway, the thing is, even if you do or don’t buy into the competing theories, the phenomena itself do appear to exist and therefore we might as well exploit it.
For computers, it was thought that you could construct the bits and bytes, the memory components, utilizing the quantum approach, and potentially be able to dramatically speed-up the computer memory capability.
Since this is a different approach to conventional computers, it has become known as quantum computing, consisting of a special computer that leverages the quantum aspects and does so using what are called qubits. Qubits are somewhat akin to classical bits but are revved up on steroids so that they are essentially faster, a heck of lot faster than today’s conventional computers.
Right now, it is extremely expensive to build a quantum computer, it is relatively large, it requires immense cooling, and it has just a smattering few qubits, thus, it isn’t anywhere near to its full potential as yet.
Of course, you can’t go around saying that computers are too big and bulky, which some said during the era of vacuum tubes and predicted that computers would always need an entire room of space, yet today we have smartphones in our pockets that are many times more powerful than the early mainframes.
It is reasonable to assume that quantum computers will ultimately get smaller, more robust, less costly, and so on, though it is going to be many years, likely decades, before that happens, if it can happen at all.
You might be wondering what a quantum computer can do for you that a classical computer cannot do.
Why care about all this chatter about quantum this or that?
Well, in theory, a quantum computer can be vastly faster at computational efforts than can a classical computer. We’re talking extraordinarily super-duper faster.
One area that this faster speed is stoking concern involves the use of encryption that we have in our digital world to protect our text messages and our files stored on faraway disk drives.
Most people assume that encryption is impossible to crack without having the secret key.
Most of today’s encryption algorithms rely on the aspect that to crack contemporary encryption would require a tremendous amount of computing power and therefore is computationally expensive or nearly intractable.
If it took the fastest conventional computer say 10,000 years to crack your encrypted files, you could suggest that for all practical purposes the encryption was uncrackable, even though the truth is that it could be cracked if you were willing to wait those ten thousand years while a computer is humming away at doing so.
Via a quantum computer, you might be able to crack an encrypted file in months, or maybe just days, or perhaps even in minutes, depending upon the type of encryption used and the size or capability of the quantum computer.
This is scary since it means that someone grabbing up encrypted files of today might sit on those files and wait until quantum computers get better situated, and then decide to crack the files and see what secrets were being hidden away at that earlier time.
The other side to that coin is that we can shift toward stronger encryption with classical computers, and likely exploit quantum computers toward forging tougher encryption schemes.
Quantum computing is one of the hottest trending fields and will continue ahead with great fanfare.
Open questions abound such as whether there are classes of problems that can be solved by quantum computers that cannot be solved at all by classical computers, meaning that other than speed alone, perhaps there are types of problems we’ve never touched or thought to solve with classical that could be shorn with quantum.
One debate too is whether there are problems that classical computers can solve that perhaps a quantum computer cannot solve or that the quantum world does worse in trying to solve in comparison to a conventional computer.
This leads us to the quantum supremacy topic.
When you get various learned computer people into a room and they have different ideas about how to do things, invariably they form into opposing camps. One camp will claim that they have a better approach, while the other camp will refute the claim.
It’s like rooting for your favorite football team or preferred brand of whiskey.
Those in the quantum computer camp had proposed that someday there would be an indication that a quantum computer could outdo a classical computer. When that day arrived, it would showcase the supremacy of quantum computers, ergo, the infamous slogan of quantum supremacy.
Google alleged that their recent effort met the quantum supremacy goal and that the task they performed with their experimental quantum computer would have taken 10,000 years for a conventional computer, but others disagreed and asserted that a classical computer could perform the same task in about 2-3 days.
One can understand the intense desire to raise the quantum supremacy flag by those that are squarely in the quantum camp, though the goalposts for measuring “supremacy” are somewhat ill-defined or some argue poorly defined (there are disputes as to what metrics to use), and some assert that it was jumping the gun to make such a bold proclamation.
Also, if you look at the zillions of problems being solved today by conventional or classical computers, those problems are not yet able to be solved by the quantum computers still being formulated in R&D mode.
Is it fair or reasonable to say that quantum is supreme when it is still far less capable than today’s everyday computing across-the-board?
Tilting the acrimony on its head, perhaps this argument about supremacy is lacking in civility anyway, and unbecoming, one might argue.
Yes, the supremacy crown is a means to spur development toward quantum computers, and yes, the esprit de corps perhaps gets the creative juices flowing and can attract public attention, but in the end, hopefully, we view computers as a tool and one that all of us ought to be finding ways to further extend and advance.
I’m sure that I’ll be getting dinged by both camps in my efforts to amend the potential polarization on this topic, which is par for the course these days (sadness).
Let’s shift our attention to true self-driving cars and how they will be impacted by the advent of quantum computers.
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 that the AI drives the car entirely on its own and there isn’t any human assistance during the driving task.
These driverless cars 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, such cars aren’t particularly going to be different per se than conventional cars when it comes to the quantum topic.
It is notable to point out that in spite of those dolts that keep posting videos of themselves falling asleep at the wheel of a Level 2 or Level 3 car, do not be misled into believing that you can take away your attention from the driving task while driving a semi-autonomous car.
You are the responsible party for the driving actions of the car, regardless of how much automation might be tossed into a Level 2 or Level 3.
Self-Driving Cars And Quantum Computers
For Level 4 and Level 5 true self-driving cars, quantum computers can be a handy aid.
First, let’s put aside the near-term chances of quantum computers becoming small enough, cheap enough, and otherwise be amendable to being used in a car as an on-board form of computer processing.
I’m knocking that out of the running for now.
Way beyond the near-term it might be feasible to have on-board quantum computers, but don’t hold your breath for it to happen.
Does that toss out quantum computers from being useful for self-driving cars?
Heck, no!
Keep in mind that self-driving cars will have OTA (Over-The-Air) electronic communications capabilities, doing so to grab system updates for the on-board AI system from the cloud, and to push data collected by the self-driving car up to the cloud.
Quantum computers could have a welcomed computationally proficient slot in the cloud for the aiding of true self-driving cars.
This is not some far off futuristic notion since today there are already cloud-based quantum computer resources available, for which researchers and others are playing with online.
Indeed, I’ve done quantum computing programming via the cloud, and it is a blast, plus it highlights that we are all still figuring out what kind of specialized programming languages ought to be used, and what kinds of specialized database-like structures should be used for quantum computing.
One caution to keep in mind is that today’s early versions of quantum computing are often fraught with high error rates. The exploration of how to contend with and mitigate or reduce system error rates in the qubits is an ongoing focus of inquiry (typically known as the quantum noise problem).
Overall, an in-the-cloud quantum computer (preferably not a simulation of one), could be used in a myriad of ways to help self-driving cars.
For example, the AI system that’s on-board the self-driving car is fed updates from time-to-time that are based on analyses of the roadway data being collected. Via forms of Machine Learning (ML) and Deep Learning (DL), the AI system is enhanced while sitting in the cloud, tested on simulated roadway data, and then when readied it is pushed out into the driverless cars for in-car use.
The computational effort of the ML/DL while in the cloud can be huge, consuming tons and tons of classical computing cycles. This means that it can take a while to craft new updates for the on-board AI system, plus it can be quite costly to chew-up all those conventional computer cycles.
Into the picture steps quantum computers.
You could have a quantum computer in the cloud that participates in the AI enhancing task, and because of the tremendous speed advantages (possibly), you might be able to generate a revised AI that can be sooner pushed out to the driverless cars in a fleet.
Think of it this way.
Suppose some driverless cars happened upon an edge or corner case of some roadway instance, one that hadn’t been earlier experienced by any of the self-driving cars in a fleet. This data might be pushed up to the cloud, crunched right away, and a new update flashed out to the driverless cars, quickly getting them all up-to-speed based on the latest learned element from other fellow self-driving cars.
I’m not suggesting that the quantum computer did something new, and instead emphasizing that we can exploit its speediness as part of the updating efforts for the driverless cars in the field.
If done properly, it means that those driverless cars are better off sooner, meaning that human passengers are better off sooner, and likewise anyone that comes near to a driverless car is presumably better off since the self-driving car has gotten an enhancement.
Don’t misinterpret this to suggest that the quantum approach is a panacea.
There could, unfortunately, be a temptation to also use the speediness to inadvertently cut corners and push out some update that isn’t ready for prime time.
Hopefully, the self-driving tech makers and those leveraging the cloud capabilities involving quantum will be as steady and somber with quantum computers as they would with everyday conventional computing.
Conclusion
Another facet of utilizing quantum computers while in-the-cloud would be to use them for doing the scheduling of car traffic and undertaking traffic management.
Some believe that once we have a prevalence of driverless cars, they could interact and coordinate to reduce traffic congestion.
Besides electronically communicating with each other via V2V (vehicle-to-vehicle) electronic communication, and via V2I (vehicle-to-infrastructure) communication, they might also interact with a “master” traffic management system that is trying to even out the flow of thousands of cars on the roadways.
The computational effort to do traffic management can be quite overbearing, so the speed of quantum computers in-the-cloud might aid such an effort.
I realize that some of you might be wondering whether the quantum computer could be fast enough that even though it is sitting in the cloud that it could nonetheless become the driver of a self-driving car.
I vote no.
As per my earlier posting that covers the dangers of using wireless as a driving approach, the risks associated with the driver being remote, whether a human remote driver or a quantum computer remotely driving, does not bode well for the safety of self-driving cars.
That being said, the on-board AI could certainly consult with the quantum computer in the cloud, doing so to get an added “opinion” about a driving matter, though this would need to be done under the expectation that there is a latency of getting a result from the cloud (or, no answer at all due to the cloud connection being untenable).
Quantum computers will be a vital addition to the portfolio of computing capabilities.
The field of quantum computing is still in its infancy.
You are now anointed as a quantum conversant (the latest “in” term for becoming familiar with quantum computing).
I encourage anyone that is eager to contribute toward the future of computers to consider getting involved in or becoming more aware of quantum computing.
I ask just one small promise.
Keep the supremacy clamoring to a modest level and do not go around wearing those Quantum Supremacy blaring T-shirts unless you are also willing to include Classical Supremacy on the same shirt with the same boldness and earnestness.
Just aiming for world peace, thanks.
