We might think that we have trained ourselves to understand the world the same way data scientists train complex AI algorithms for predictive purposes, but that's not always the ground truth.
We usually like to think that we know every little detail about ourselves, but in fact there's so much to discover. I might see myself as this quirky wallflower from Eastern Europe, yet others might be fed up with my occasional in-your-face attitude. It's not only about how others perceive certain people. Look at your 4th-grade-self, for example. Those were some bold decisions, and I'm not talking just about those bell-bottoms. It's all about the perception of things, people and yourself at certain points in time. We might think that we have trained ourselves to understand the world the same way data scientists train complex AI algorithms for predictive purposes, but that's not always the ground truth.
Everybody sees themselves in a different way. I am well aware of my physical shortcomings and I'm highly self conscious about them. Ask someone about your own insecurities and you'll actually be surprised that people don't really notice the things you are obsessed with. We focus on our flaws, but not everyone sees them and cares in the first place. You think you know who you are, but there are actually multiple versions of you. I'll explain.
Your perception of self changes as you get older. When I was a boy that was the only version of my person. I knew that I enjoyed playing soccer with my friends and I absolutely despised mashed potatoes. That was my only reality. When I'll be an old man, I will find joy in my life in completely different ways than I did as a kid or now. Both of these people are strangers to my current self, but essentially both of them are me.
Humans are complicated creatures, and what makes us exactly who we are is hidden deep beneath the surface. Even throughout the day, you are not always the same person. As we interact with various people we reveal different layers of our emotions, thoughts, and motives. I'm positive you talk to your grandma in a very special way. You would never snap at your mother the same way you would at a waiter that messed up your order. Therefore is there even such thing as your true self? What is the ground truth?
Ground truth could also be called just truth. Reality. Ground truth is also the bedrock of AI and deep learning. It's the foundation of any algorithm, any model to successfully categorize visual data or predict the potential outcomes of a certain event in any industry. Accurate and successful object detection in autonomous vehicles is one of the most challenging and difficult features to achieve full autonomy. Self-driving vehicles need to be able to distinguish between other cars and pedestrians to safely navigate in any environment.
It's a well-known fact that some sensors are still too expensive to massively penetrate the market, detection range might not be long enough and these types of vehicles are highly vulnerable to cyber attacks. I'd like to mention the perception of the environment as one of the main cornerstones for full deployment of self-driving cars. Simply detecting and observing certain objects don’t really make the vehicle navigate around streets, obstacles, or people. It has to know exactly what those objects are, which is why the car’s ability to identify a cyclist versus a child is so important. Fundamentally, cyclists and kids on the streets behave differently and understanding their general behaviors can help the vehicle reliably predict what might happen next.
Today's predictive models and algorithms are predisposed to misunderstanding and false assumptions if they are not fed with enormous amounts of high-quality data.
Today's predictive models and algorithms that give intelligent machines the ability to analyze data are predisposed to misunderstanding and false assumptions if they are not fed with enormous amounts of high-quality data about the real world. It is an oversight to theorize before and conclude anything before we have data, and that applies to virtually any situation. We're usually quick to see the faults in others, while for the most time we're unconscious of our own. We sometimes assume things about others just based on brief interactions or just because we were fed up with their actions that one time. It's always necessary to see the entire picture from multiple perspectives in both self driving scenarios and our daily lives.
HD mapping platforms, as well as a wide variety of sensors, are being used to perceive the entire situation around the self-driving car. Only when all these technologies are tied together, we can hope for a high level of safety and reliability of these vehicles.
For autonomous cars, there is no room for mistakes or errors. Theoretically, self-driving cars will reduce the chance of crashes and fatalities by perceiving the environment around them and efficiently communicating with other cars and the surrounding infrastructure. However, the safety argument is respectable only as long as AI and the algorithms behind the vehicle are reliable. Things that are easily understandable to a human driver can be very confusing for an autonomous vehicle. Tree shadows, stickers on street signs, snow and raindrops - all of these things create difficulties for sensors to understand the environment. nuTonomy even had to reprogram their algorithms because of some stubborn seagulls casually walking on the streets and not being bothered by the self-driving quiet electric vehicles approaching them.
As I mentioned earlier in the article, introverts and perfectionists often criticize themselves and seek approval from the people around them. This need for validation continues into adulthood - people are trying to only show their best selves to be validated by their peers, friends, family. Eventually, everybody grows the wisdom to realize that mistakes and failures are a part of the human experience. We as humans learn from our mistakes the same way as AI algorithms do, except we can't afford vehicle perception models train themselves based on real-life situations only. That's why the automotive industry is trying to optimize self-driving car development and testing process with simulation.
Simulations allow experimenting with a multitude of changing conditions to prepare autonomous vehicles for their full deployment in the future. Sure, it's a great way to fabricate various scenarios and add theoretical knowledge to the brain of the vehicle, but it's not exactly real nor accurate. We can read all the self-help books in the world and familiarize ourselves with psychological research, but the true lessons are learned through real-life experiences. I believe that is the best way to stay true to yourself and maintain the authenticity of your personality, even though hiding behind Instagram filters and Facetune sometimes is way easier.
Why are we so afraid to be authentic? Our imperfections make us human, unique and relatable. We all have moments when our troubles feel big and our faith feels so small. Let those who love you carry you. If you met yourself would you like the person? If you were hesitant even for a tiny second, something is wrong.
The majority of us can clearly tell when we are in the presence of an authentic and true human being, one who is not trying to be someone they are not, they are just true to their humanness. This kind of comfort can be felt because in their presence we can evidently sense our own authenticity and we sense the deep peace this authenticity brings. Same goes with self-driving cars - only when we will feel beyond doubt confident about their technical readiness, this wild and life-changing idea will become a fully legitimate reality.
Our journey from childhood to adulthood and the self-exploration aspect of that is very similar to the development process of any modern technology in the form of multiple iterations and new advancements. The autonomous vehicles are still in their infancy, but it doesn't mean we have to be too. People will always perceive things differently based on their past experience, but as long as we stay true to ourselves and put things in perspective before any judgments are made, there is still hope for a better world tomorrow.