Human+AI: the key to success?
If you ask our CEO Patrick Vanbrabandt when he believes that we will arrive at Level 5 full driving automation, you could probably expect an answer along the lines of “We need to adjust our expectations in the matter.” This is what he means by that:
Technology has already evolved a long way when it comes to transportation. We have arrived at the point where cars are for about 90% self-driving but still need humans for that last 10%. You can compare it to flying an aircraft. Most of the journey now happens on auto-pilot. Should there be a problem with the human pilot, chances are the plane will be able to safely land because of the technology integrated in the airplane. It’s about the same with cars: they can park by themselves, drive on cruise control, brake for sudden obstacles, notify the driver when he’s getting drowsy and make all kinds of autonomous and helpful decisions. There is so much intelligence included in our cars already.
And yet, people seem pretty much obsessed by that last 10%. But is it so bad that humans keep some small portion of control? Take BMW’s approach for instance. Their latest models only autonomously steer to the right if the driver checks his right rear mirror, even if he’s not in charge of driving himself. Yes, this is about safety, but there is also a psychological aspect to it: it’s about peace of mind, about reassuring our brains that nothing is going to happen.
So perhaps, rather than expecting (self-driving) technology to be perfect all the time, we should adjust our expectations as well as our definitions. Maybe we should aspire towards hybrid - human plus AI - driving rather than 100% autonomous driving.
We could compare it to Generative AI systems like ChatGPT and Gemini. People seem to accept the fact that these are imperfect systems that make up things. In the field of AI, these types of responses which contain false or misleading information are called “hallucinations” and they happen quite frequently. That is why no one should ever let them act without any supervision. Yes, we expect them to gain higher independence along the way, when autonomous agents have matured, but we’ll probably need to do one last human check on them for a very long time. In fact, mayor tech brands have already admitted that they’re even not sure that they’ll be able to ever control these “hallucinations”.
So let’s use the same approach here, as with other technologies and go for a human-AI hybrid form where both reinforce and augment each other. Where the computers take care of the boring, repetitive stuff and the humans perform the exciting, creative tasks and exceptions. Not only will this be safer and way more fun, we will also make sure that our human value is never taken out of the equation.
Want to learn how CARYA can help your company with a hybrid human-tech approach that will benefit your customers? Check our website to see if we can help you with your CRM, DMS, IT infrastructure or marketing!
And yet, people seem pretty much obsessed by that last 10%. But is it so bad that humans keep some small portion of control? Take BMW’s approach for instance. Their latest models only autonomously steer to the right if the driver checks his right rear mirror, even if he’s not in charge of driving himself. Yes, this is about safety, but there is also a psychological aspect to it: it’s about peace of mind, about reassuring our brains that nothing is going to happen.
So perhaps, rather than expecting (self-driving) technology to be perfect all the time, we should adjust our expectations as well as our definitions. Maybe we should aspire towards hybrid - human plus AI - driving rather than 100% autonomous driving.
We could compare it to Generative AI systems like ChatGPT and Gemini. People seem to accept the fact that these are imperfect systems that make up things. In the field of AI, these types of responses which contain false or misleading information are called “hallucinations” and they happen quite frequently. That is why no one should ever let them act without any supervision. Yes, we expect them to gain higher independence along the way, when autonomous agents have matured, but we’ll probably need to do one last human check on them for a very long time. In fact, mayor tech brands have already admitted that they’re even not sure that they’ll be able to ever control these “hallucinations”.
So let’s use the same approach here, as with other technologies and go for a human-AI hybrid form where both reinforce and augment each other. Where the computers take care of the boring, repetitive stuff and the humans perform the exciting, creative tasks and exceptions. Not only will this be safer and way more fun, we will also make sure that our human value is never taken out of the equation.
Want to learn how CARYA can help your company with a hybrid human-tech approach that will benefit your customers? Check our website to see if we can help you with your CRM, DMS, IT infrastructure or marketing!
