laitimes

Musk, who has threatened to achieve full self-driving within the year, has punched the face four times this year?

On April 18, Musk made another bold statement, saying that Tesla will achieve full self-driving within this year. At first glance, this rhetoric sounds familiar, yes! In 2015, 2017, 2020 and 2021, he made the same rhetoric four times, at different times, with the same result, and was punched in the face, which was the fifth time. For others, this kind of wolf's behavior three times is the ceiling, and more is annoying, but this is Musk, even if he punches the face four times, the fifth time is still eye-catching, because he is not an ordinary guy!

Musk, who has threatened to achieve full self-driving within the year, has punched the face four times this year?

Why you punched your face four times is still a matter of concern

First of all, looking at financial resources, Musk once became the world's richest man with a value of 1,280 billion yuan. Looking at his achievements, he has made remarkable achievements in all aspects, whether it is in the heavens, into the earth, science and technology, or in the humanities.

Musk owns Spacex Space Exploration Technology, which can realize rocket recycling and greatly reduce the cost of space transportation, and has now achieved commercialization. In addition, the company's next goal is to achieve human interstellar navigation and strive to create a true space civilization for human beings. He is also the founder of PayPal (the world version of Alipay) and the originator of mobile payment; he led another company of his own, Neuralink, to pioneer human-computer connection technology, through the machine to connect the brain to achieve the autonomous control of human brain consciousness Machine has achieved initial success with pigs.

Musk, who has threatened to achieve full self-driving within the year, has punched the face four times this year?

Recently, he just bought Twitter for $44 billion and threatened to buy Coca-Cola; this year he will build a hyperloop through his "Boring Company" (The Boring Company), breaking through 1,000 kilometers per hour, faster than the Boeing 747, announcing that he expects to start full testing this year in 2022; he also said that he will start producing a humanoid robot called Optimus as early as next year, that is, 2023...

It can be seen that his near-term goal is not only to achieve fully autonomous driving. Someone else might be mistaken for a lunatic, but this is Musk, a man who has accomplished too many seemingly crazy goals, so even though his self-driving career has been punched in the face four times, it still catches the eye when he threatens his fifth time.

Can the fifth autopilot flag be achieved?

First, let's take a look at what level Tesla's self-driving technology is at as a whole in the field of autonomous driving. Tesla began to sell FSD (Full Self Drive), that is, full self-driving software, which is still in a semi-finished state, equivalent to an enhanced version of autopilot assisted driving, in some sections of the road can independently complete deceleration, lane change, turning, courtesy to pedestrians and so on. As far as users are concerned, its overall performance is not bad, but the ranking of the entire autonomous driving field is not ideal.

Musk, who has threatened to achieve full self-driving within the year, has punched the face four times this year?

For example, Navigant Research conducts a comprehensive evaluation of 15 autonomous driving companies around the world from ten dimensions of corporate vision, marketing strategy, partners, production strategy, technology, sales & marketing & distribution, degree of commercialization, R&D progress, product portfolio and financial strength. At the same time, according to the execution ability (Execution) and the strategic ability (Strategy), these players are divided into four levels: "leader", "competitor", "challenger" and "follower". In the international autopilot rankings, Baidu Apollo continues to be in the "leader" camp, while Tesla has been at the bottom in recent years.

And Tesla has so far more than two hundred autonomous driving-related safety accidents, including out of control, collisions and other major accidents, although most of the accidents are caused by drivers relying too much on the ability of FSD, but this also shows that FSD is still in a semi-finished state. Just last August, the National Highway Traffic Safety Administration launched an investigation into Tesla autopilot, involving 765,000 cars, and from this perspective, it seems that Musk's desire to achieve full self-driving this year is still unlikely.

Musk, who has threatened to achieve full self-driving within the year, has punched the face four times this year?

But from another point of view, although Tesla's automatic driving comprehensive ranking is at the bottom, Tesla's strength and future should not be underestimated. Tesla develops autonomous driving based on vision solutions, and the current neural network architecture based on pure vision solutions has been basically built, and 80% of the current research and development energy is doing data training. Tesla released the FSD Beta (fully autonomous driving beta version) in October 2020, and took the lead in the United States market promotion, so far the version has been updated dozens of times and maintained a high frequency of updates, optional installation of the function and driving safety score to a certain standard of users can use the function, in the use of tesla is also to help Tesla train the ability of FSD, until the FSD processing power exceeds that of humans.

Musk, who has threatened to achieve full self-driving within the year, has punched the face four times this year?

In a few years, FSD Beta has expanded from a limited number of 20 people to 100,000 people around the world, so its efficiency in obtaining data is extremely impressive. Although it seems that users are suspected of taking users as guinea pigs, on the one hand, this is a wishful thinking, on the other hand, such a large user volume for data training is much more efficient than other self-built teams of automatic driving companies, so the rapid progress of Tesla's autopilot technology and even the ranking counterattack is still very promising, and it is not completely impossible to achieve full automatic driving this year.

Even if the technology is achieved, landing is still a problem

At present, the laws and regulations for automatic driving are not sound, and many countries have just recognized the L2 level of assisted driving, while full automatic driving belongs to the L5 level, and such a span is difficult to be accepted by the regulations within one year. In addition, if you want to fully land on automatic driving, there is a particularly difficult problem - the accountability mechanism is not clear. Traditional human driving to L5 level before the automatic driving, there are human participation, once there is an accident, to the driver to pursue responsibility is reasonable, but full automatic driving does not require driver participation, once there is an accident, whether it is to "over-trust the driver's seat members of the automatic driving system" accountability, or to the accident car manufacturers accountability, it seems unreasonable, before this accountability problem is clear, automatic driving technology even if it is developed to a certain extent can not be achieved.

Resistance on many fronts

In addition, Tesla faces resistance not only to regulations and accountability issues, but also from competitors and users. For example, American billionaire Dan O'Dowd publicly stated that the beta version of Tesla's FSD was "terrible" and said that he would launch a multimillion-dollar advertising campaign in the United States, calling for a "ban" on Tesla's fully autonomous driving (FSD) optional package.

Musk, who has threatened to achieve full self-driving within the year, has punched the face four times this year?

The rich man also deliberately ran for the Senate in order to block the FSD software. It is understood that the rich man also has companies engaged in the field of automatic driving, so he is both a competitor and a typical representative of citizens' resistance to automatic driving technology, after all, it is not easy for many people to completely hand over their lives to the hands of software.

It can be seen that it is not easy to achieve fully autonomous driving this year, and Musk is likely to complete the fifth punch in the face this year. But one seemingly impossible thing after another has been done by Musk, and in comparison, the difficulty of getting autonomous driving to land during the year is not so great for Musk.

Read on