AI That Tries To Be Funny Is Not Necessarily A Laughing Matter, Especially When Used By Self-Driving Cars


Trying to infuse AI with humor telling capacities is a challenge, plus we might wish to be mindful … [+] of comedic AI in systems such as self-driving cars.


Humor can be quite useful.

Let’s see how.

Suppose you are having a bad day (I realize this seems a bit dour and gloomy, but the venerated gallantry of well-placed humor will turn this around, wait and see).  

While on the way home from work, you opt to use a ridesharing or ride-hailing service. After settling into the car and getting ready to have a quiet and solemnly introspective ride home, the driver suddenly speaks up and tells you an entertaining joke that causes you to laugh out loud. The merriment of the humorous anecdote sparks you to get out of your dismally sour mood. It was one of the best jokes you have ever heard, a knee-slapping and side-splitting piece of humor.

Things are looking up, and when you arrive home, your perspective and attitude about the world have shifted into one of peacefulness and content.

All due to letting some humor enter into your day.

Admittedly, that is the smiley face version of humor. There is also the other side of the coin, namely humor that is off-target, bleak, and at times insulting and entirely offensive.

You might have gotten into that ridesharing vehicle and the joke told by the driver could have inadvertently made your day even worse. How dare the driver tell such a joke? It was disgusting and abundantly inappropriate.

The thing is, the situation or contextual setting under which humor is encountered can make a big difference in both how the humor is received and what you can do about the presented humor. For example, when watching a TV or cable show that is spewing forth foul humor, you can presumably switch channels or turn off the offending drivel. If you are at work and your boss tells you a mildly improper joke, you need to weigh the potential consequences of telling them that the humor was misplaced versus simply doing a weak smile and letting the matter slide past you.

In short, your ability to react to humor can be mitigated by the circumstances under which the humor is being employed.

Return to the use case of being a passenger inside a car. When the driver tells an altogether untoward joke, it might seem obvious that you should right away correct the driver and tell them they have transgressed. On the other hand, you must also weigh the fact that the driver is the driver. This means that the fate of you and the vehicle is in the hands of that driver. Any immediate and debate-sparking response to the driver about a bad joke that they have just told could rattle them and the next thing you know the car is veering or being driven recklessly. One supposes that you are inclined to laugh at any humorous utterances of the driver that are within bounds, merely to placate them and try to keep attention to the act of driving versus a prolonged and driver-distracting discussion about the appropriateness of joke-telling.

Okay, so we’ve established that there can be adverse consequences when foul humor is used, including rancorous consequences for the person telling the joke and for the person receiving the humor. Meanwhile, by luck of the draw, humor that turns out to be well-received can be an upside for both parties, though this has to be weighed against the potential downsides that can also be incurred.

There is a tradeoff involved in leveraging humor, that’s for sure.

Should someone try to use humor and aim for the aspirational goal of providing humor that brightens the day, or avoid humor due to the potential for getting into the muck and ruining things for all involved?

A ridesharing driver might believe that telling jokes is a great way to bond with passengers and ostensibly get a better rating or earn a heightened tip for doing the driving. Of course, not everyone is made of the right stuff to be a comedian. The joke teller might provide a joke that falls flat and seems to be entirely bereft of anything funny at all. The attempt at comedy could go beyond simply being flat, and venture into the zone of being offensive.

Most ridesharing services seek customer feedback from riders. A joke-telling driver that routinely gushes out lousy humor that rankles passengers is likely to end up with low ratings and risks losing their privileges of being a ridesharing driver. In that manner, some passengers might decide to quietly or silently ding the joke-telling driver via scoring them in an app, rather than confronting the driver directly during the driving journey.

Another possibility of letting the driver know that the humor was humorless and distasteful would be to wait until the ride is over.  After exiting from the vehicle, you can now relatively safely unleash your tongue and provide the driver with a bit of some lesson-gleaning insights about how to not tell jokes. Some drivers will get the drift, while others will shrug off the feedback and figure that they’ll continue to use their humor on other riders and write off your reaction as an outlier reaction.

Scholars that have studied humor tend to divide humorous utterances into one of four types, labeled as affiliative, self-enhancing, aggressive, and self-defeating.

Not everyone necessarily agrees that those are the complete and exhaustive list of humor types, and there is even disagreement about the scope and slippage amongst the claimed four types. As they say, your mileage may vary (wait for a second, was that a kind-of joke or humorous anecdote about the nature of humor?).

We can go ahead and address each of the four types and then think about the ramifications thereof.

Affiliative. The affiliative type of humor is used by the joke teller to try and bond with others and create an affiliation, giving rise to a social relationship with the receiver.

Consider this in a driving and rider context.

The driver of a ridesharing vehicle doesn’t know you and you don’t know them. A handy ice breaker or opener to the conversation could be a well-timed and well-shaped piece of humor. That might start the ball rolling in terms of chatting throughout the traversal. Notably, it could also stomp down on any chance of conversation and be so off-putting that only stony silence remains the rest of the ride.

Self-enhancing. Self-enhancing humor tends to involve telling jokes that will enhance the image or semblance of the teller, doing so in the eyes and perspective of the receiver.

In a driving journey context, a driver might make a joke as an attempt to demonstrate that they are more than simply somebody driving a car. The rider suddenly realizes that the driver is a human being, a regular person, and not some unfeeling form of automata that perchance is at the wheel of the vehicle. Humor in this context is intended to showcase something about the joke teller. It might cause the receiver to think of the teller as being clever, witty, smart, etc.

The backlash can be that the receiver perceives the joke teller as unrefined, rude, outlandish, out-of-touch, or suffers from other similar societally undesirable traits.

Aggressive. The aggressive type of humor seeks to enhance the person telling the joke and yet do so at the expense of the person receiving the joke. That’s a dicey line to hoe. Perhaps you’ve seen the antics of a Don Rickles or akin type of comedian.

For a driver, imagine that they opt to make a joke about what the rider is wearing, such as a hat or coat, or shoes. The receiver of the humor is being made fun of. This can right away go awry. Perhaps the worn apparel is beloved. Any jokes about the clothing could be instantaneously rebuffed.

There is a chance that the humor might work. Maybe the rider has always thought that their hat was a bit silly and pretentious. The driver pointing out this facet might prod the rider into a fit of laughter.  Good luck on trying those potential landmines.

Self-defeating. The self-defeating type of humor is the now-classic and widely utilized self-deprecating or self-disparaging approach in our daily narrative, laced with something funny mixed into the concoction.

Envision a driver that tells you something humorous about themselves that is actually self-disparaging. I used to be a medical doctor, the driver might say, but I decided that there wasn’t enough stress in saving people’s lives in the operating room and ergo chose instead to relish an even more stressful life by driving in the congested traffic of this great city.

Maybe you find that funny, maybe not, but the point is that the humor aims to undercut the teller and as being told by the teller. The receiver is often leaned toward sympathizing with the joke teller due to the self-deprecating aspect of the humor. It seems as though the teller is proffering a sense of humility, though we also know these days that people on online social media try to use this approach as a form of honor brag.

When AI Becomes Humorous

This discussion has brought to the surface and highlighted that humor is a dual-edged sword.

Humans sometimes get things right and tell a good joke at the right time to the right receiver. Presumably, that’s how professional comedians make their money, somehow knowing how to strike that chord.

The problem with a dual-edge sword is that you can get pretty banged up and harmed by the other edge of that slicing and dicing. Thus, humans sometimes get things wrong and tell a bad joke at the wrong time and to the wrong receiver. Once the joke has been told, the chances of unwinding it are relatively slim. You can try to backtrack and assert that you didn’t mean it. You can try to tell another joke that tries to cover up the offensiveness of the already uttered joke. The problem is that once the bell has been rung, unringing the bell is a huge challenge.

Let’s slightly shift gears and go beyond the realm of human joke-telling.

What about Artificial Intelligence (AI) as a joke teller?

Believe it or not, some believe we ought to have AI systems that can utilize humor.

You can plainly see the logic by considering what the overarching aim of AI espouses to be. The viewpoint is that AI is an attempt to devise machines that can exhibit intelligent behavior. By and large, this aspiration is shaped around making computer systems that seem to act smartly or intelligently, akin to what humans do or are supposed to do.

Upon giving some reflective thought about human behavior, we can likely all agree that humor seems to be a component of human behavior. Going further, some would argue vehemently that humor is a vital or integral element of human behavior. In brief, perhaps the notion of intelligence is also wrapped into the inoculation of humor. The assertion is that humor is either a needed aspect for giving rise to intelligence, or it is at least an integral byproduct that emerges from intelligence.

You might not buy into that theory.

The counterargument is that there is very little to do between intelligence and humor per se. We can have perfectly fine intelligence and have absolutely no humor somehow infused at all. Humor is a completely separable characteristic. In fact, by daring to contend that humor is somehow essential to intelligence, you are going to distract the strident pursuit of crafting machine-based intelligence and divert attention into something that has little or no substantive bearing on the genuine matter at hand.

I’ll let you mull over that philosophical conundrum.

Meanwhile, there is not much doubt that today’s AI systems could potentially make use of humor and indeed this is already happening. Whether this gets us any closer to the far reaches of attaining AI is a separate, interesting, and discernable question, but not one to be solved at this moment.

You might have interacted with an AI system that purportedly is using humor. These are usually pre-canned humorous statements that the programmers planted into the system and are nothing more than the recital of text. The AI is not sentient. It has no “understanding” about the humor. The jokes being told are even less than a monkey-see-monkey-do since the odds are that a monkey would potentially have a grasp of the meaning of humor (there is interesting research about the use of humor by animals, and you might find it worthwhile to take a look at such works).

The use of humor in an AI system could be classified as being flat or of a recital nature and essentially be entirely prescribed beforehand by the AI system designers and developers. That is about the least or lowest level of using humor in an AI system.

A step up from there entails AI that has been crafted to try and create humor on the fly. This is a lot trickier. Further up the ladder would be AI that is both able to emit humor and receive the humor, doing so in an interactive mode.

Where would AI that can emit humor be used?

The answer might seem smarmy but there is an opportunity to use humor in AI wherever the AI is being used that otherwise, a human might have opted to use humor.

That might seem confusing.

To clarify, suppose you go to see your primary physician about a sore shoulder. You raise your arm and explain to the medical doctor that it hurts when you do that. The physician looks at you and in a straight-faced voice declares, well, don’t do that (bada-bing bada-boom!). This rather worn-torn joke showcases that humor can be used in the medical field. As repeatedly emphasized earlier, leveraging humor can be a stress reducer and be tremendously helpful or it can be poorly pitched and be abysmal. The same applies to using humor in medicine.

There is another realm that AI and humor can be considered for use. All this talk about ridesharing and what a driver might say is a handy segue into the topic of AI and self-driving cars (for my extensive coverage about AVs and self-driving cars, see the link here).

The question to consider is this: Should AI-based true self-driving cars make use of humor?

Time to unpack the matter and see.

Understanding The Levels Of Self-Driving Cars

As a clarification, 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 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-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 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 Using Humor

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 (see my analysis at this link here).

Why 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 humor and the use of humor. Any use of humor and whatever 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.

You get into a self-driving car and the AI driving system greets you with a joke.

Is that a good thing to do or a bad thing to do?

Well, if you are stressed out about potentially using a self-driving car, the use of humor could put you at ease. You might relax and feel like you are simply going on a ride, as though a human was driving the vehicle.

On the other hand, you might interpret this use of humor as horribly misplaced. You want an AI driving system that takes the act of driving as a quite serious matter. Maybe the AI doesn’t care much about the troubles of driving and instead the focus is on joke-telling. When your life is about to be placed into the hands of the AI, you’d rather that it was attentive to the roadway.

There’s that double-edge sword again.

Most of the self-driving car tryouts are likely to eschew using humor for now. Keep things simpler and cleaner, for the moment, and maybe add some touches of humor later on.

A complaint that is likely to get lodged about AI using humor in the case of driving a self-driving car is that it could mislead passengers into believing that the AI is in fact sentient. This takes us back to the discussion about navigating away from portraying AI as though it is on par with humans (we ought to avoid going the anthropomorphic route).

An alternative way to sneak humor into this realm would be to have the humor delivered by humans. You get into a self-driving car and a screen inside the car begins to show you a short video. The video shows a spokesperson that explains the greatness of this self-driving car and the aspect that you are riding in it is quite astute on your part. The recorded video then provides some humor to get you relaxed about your ride.

Notice that this keeps the humor separated from the AI. The joke-telling arose via the human spokesperson and not due to the AI system.

Here’s another twist for you.

Suppose the AI driving system does have some kind of humor-emitting component. This brings up a technical concern. If the AI driving system is using its computer processors to do joke-telling, doesn’t that imply that those same processing cycles are not being used toward the driving task? And, if so, wouldn’t it be more prudent to use any and all processing capacity to try and maximize the safest possible driving journey, rather than expending such effort on humor?

You could make that argument.

The retort or counterargument could be that the processing of the AI-based humor emission is coming from a separate onboard computer that is used for other facets such as in-car entertainment. You wouldn’t presumably be arguing that the radio or in-car TV is somehow usurping the processing for the AI driving system, so why should it be of concern to do so when telling jokes?

The AI joke-telling could be entirely distinct from the AI driving system. This might be an add-on or some other augmented feature that the self-driving car can download and use. Again, think of this as a form of in-car entertainment.

During the driving journey, the AI joke-telling might be tapping into the AI driving system to keep track of how the driving is coming along. Suppose the AI driving system takes a turn that is particularly sharp and maybe slightly jostles the passenger inside the self-driving car. The AI joke-telling system might be on alert for such situations and blurt out that it was quite a fancy turn, making light of the driving action.

Good idea or bad idea?

It might ease the tension and the rider would realize that the AI realizes that the turn was a bit tight. Or it might heighten the tension if the rider didn’t notice the sharp turning action and now, all of a sudden, become leery of what the AI driving system is doing.

One seemingly irrefutable advantage of the AI-based humor telling would be that the AI won’t get ruffled or upset if you decide the joke wasn’t funny. You could presumably berate the AI to your heart’s content, and not worry that the AI is going to change how it is driving such that it suddenly gets emotional and drives oddly. Unlike a human driver, the AI won’t have such reactive qualities (though, just to clarify, this is based on the assumption that the AI would be programmed to not do so, and that it would be just as easy to have the programmed AI to indeed become erratic or respond poorly if that’s what seemed like a desirable outcome as devised by the programmers).


The automakers and self-driving tech firms have their hands full right now as they try to get self-driving cars to simply drive from any given point A to point B.

Thoughts about adding humor are at best a lowly ranked edge or corner case, meaning that it sits on a list of things to do and resides at a very low priority. That being said, it would be relatively easy to plug in an AI-based humor component if all that you wanted to do was have some kind of AI-related humor telling capacity.

Furthermore, most self-driving cars will have inward-facing cameras and audio microphones, allowing this kind of AI to try and detect what reactions passengers have to the emitted humor. A real-time AI-based humor component of sophistication would be able to do an immediate analysis of the passengers using facial recognition and other Machine Learning pattern matching to try and discern how someone is responding to the humor. This is somewhat akin to how any joke teller tries to read the crowd, as it were.

Perhaps the next time that you feel a strident need to enjoy some peppy standup comedy, you’ll beckon a self-driving car and go for a ride. That reminds me, have you ever heard the one about the self-driving car and the self-driving truck that walk into a bar?

It is quite a hilarious knee slapper.

Leave a Reply

Your email address will not be published. Required fields are marked *


OKC Thunder Rookie Josh Giddey Continues To Exceed Expectations


If You Loved ‘Squid Game’ Here Are 5 Fabulous Korean Netflix Series