YouTuber “Wham Baam Teslacam” has shared one Tesla owner’s incredible footage of something that’s usually seen in the movies — an insane police pursuit and takedown. The Model 3 owner, Ezekiel, was driving on the highway in Oregon when he noticed the police car speeding up behind him. Ezekiel moved over to let the police officer speed by and he noticed a car in the distance coming straight toward them. The police officer lined up with the vehicle to take it down by hitting the vehicle directly. The car, which was speeding down the highway in the wrong direction, was stopped.
Immediately after stopping the other vehicle, the police officer broke the window of the other car and arrested the driver in no time. The official story said that close to 2:00 PM, emergency dispatchers were notified about a wrong-way driver on the highway near Milepost 342.
The report cited a sedan that was driving westbound in the eastbound lanes. Oregan State Police officers responded to the area in hopes of spotting the driver. Less than five minutes later, one of the officers spotted the vehicle right before colliding with it. The wrong-way driver only had minor injuries and was taken to the hospital for treatment.
After being treated at the hospital, the driver was booked in the county jail on charges of second-degree attempted assault and reckless driving, and reckless endangerment.
“Imagine putting yourself in this kind of danger to stop an individual from hurting others. Wham bam, that trooper is a hero, man,” the narrator said. The video goes on to show other Tesla cam clips sent in by Tesla owners.
Reckless Driving Kills, & Solving Real World AI Will Help Save Lives
In 2018, the National Highway Traffic Safety Administration (NHSA) reported 51,490 fatal car crashes and noted that reckless driving caused the majority of those deaths. 17% of those fatalities were caused by driving too fast for the conditions. For over 20 years, speeding has been involved in at least one-third of all motor vehicle fatalities and is at the top of the list of related factors for drivers involved in fatal crashes. The Insurance Institute for Highway Safety (IIHS) found that rising state speed limits over the past 25 years have cost almost 37,000 lives.
With autonomous driving, these problems can be solved once Tesla and others master Level 5. Many believe Tesla is on the brink of solving real-world AI with Dojo and a massive fleet. The YouTuber “The Future Economy” explains in more detail. He referred back to Tesla’s Q1 2021 earnings call, where CEO Elon Musk said the following:
“Then with regard to Full Self-Driving, Full Self-Driving Beta continues to make great progress. It is definitely one of the — I think one of the hardest technical problems that exists, that’s maybe ever existed. And really, in order to solve it, we basically need to solve a pretty significant part of artificial intelligence, specifically real-world artificial intelligence. And that sort of AI, the neural net needs to be compressed into a fairly small computer, a very efficient computer that was designed, but nonetheless, a small computer that’s using on the order of 70 or 80 watts. So this is a much harder problem than if you were you, say, 10,000 computers in a server room or something like that.
“This has got to fit into a smaller brain. And this — I think with the elimination of radar, we’re finally getting rid of one of the last crutches. Radar was really — it was making up for some of the shortfalls of vision, but this is not good. You actually just need vision to work.
“And when your vision works, it works better than the best human because it’s like having eight cameras, it’s like having eyes in the back of your head, beside your head and has three eyes of different focal distances looking forward. This is — and processing it at a speed that is superhuman. There’s no question in my mind that with a pure vision solution, we can make a car that is dramatically safer than the average person. So — but it is a hard problem because we are actually solving something quite fundamental about artificial intelligence, where we basically have to solve real-world vision AI.”
Further breaking this down, the host of The Future Economy pointed out that real-world AI is the incredibly hard problem of making a robot that runs on a neural network that is similar to how our brains work.
“Our brains are basically neural networks, too, and we have two cameras that we use to drive. So Tesla’s mission is to basically have this come to life. A Model 3 with Full Self Driving isn’t just a car. It’s technically a robot in the shape of a car. The car has an artificial brain and it uses that to navigate around the world, read road signs, avoid obstacles, and everything that a human would do.”
Additionally, the computer cars don’t get emotional. The cars don’t get angry or take offense to someone cutting off in front of them. They just react in as safe a manner as possible. The car also doesn’t panic or get scared when having to slow down suddenly or move out of the way of an unexpected vehicle that is on a sudden collision course with it. This is how real-world AI can save lives — by utilizing defensive driving tactics while also not being a part of the problem. An AI doesn’t need to text its boyfriend while putting on lipstick while making a left turn.
The above video is further broken down into three sections.
1. The fundamental problems that Tesla needs to solve Full Self Driving. There are a total of three major problems that Tesla needs to solve to make FSD work.
- Collect millions of hours of training data.
- Processing all of that data. Tesla’s solution to this is the Dojo supercomputer.
- The finished neural network has to be able to run on the FSD chip that’s in all of the Tesla cars.
2. Scaling AI today. The way most neural networks are developed is by putting them in a simulation and training them for hundreds of thousands of hours. In Tesla’s case, since it’s controlling the simulation, it can make copies of the AI and train them all at the same time. “Tesla’s approach is probably the hardest to replicate but it’s also the best shot at solving full autonomy using only computer vision.”
3. The concept of how humans solve problems and why Tesla’s AI will be so powerful. In this section, he covers the Iceberg Theory — also known as the Iceberg Principle — which suggests that we can’t see or detect most of a situation’s data. It’s also known as the theory of omission.
Another way of looking at this theory is looking at the iceberg itself. What you see is how to drive a car or to ride a bike. What you don’t see are the harder parts such as maintaining balance on the bike or memorizing the rules of the road — this part is what your brain had to learn.
The three sections above are just the nutshell version of the second video and I encourage you to watch the full video for more details. Although this second half may not seem to have anything to do with the first half of this article, they are connected. Real-world AI will someday prevent scenarios such as those presented in the first video from happening as often as they currently do.