laitimes

Jidu Battle: The chess eye of Baidu's ten-year AI chess game

author:Talk about AI

Text: Zheng Driving @ Tan Qing said AI editor-in-chief

More than 400 days after jidu built the car, the first "pre-answer sheet" has been handed over.

Although Jidu, as a joint venture brand between Geely and Baidu, also occupies 45% of the shares, it is not difficult to find that Whether IT is the conference held in Greece, or the blessing of Baidu's full-stack autonomous driving capabilities, or the participation of this conference, the latter obviously has a more positive attitude.

This point is actually very easy to understand, after all, the strategic importance of Jidu to Baidu is self-evident, Baidu has been brewing for more than ten years of AI and autonomous driving capabilities have not been greatly demonstrated in the market, in the face of the urgency of technical mass production, Jidu is undoubtedly the most important bottom card in Baidu's hands.

Jidu Battle: The chess eye of Baidu's ten-year AI chess game

It is undeniable that in just over 400 days, the first shot of this Internet giant car manufacturing did sound earlier than many people expected, and Jidu as a new player, in the competition of the white-hot car-making army, the wild hope of the car robot does have some head iron.

It is difficult to avoid that after the unveiling of ROBO-01, there are also doubts wrapped up with many surprises, which are either from the viewers in the live comment area, or from li xiang, Guo Tiefu and other industry people.

But this car robot, he finally came with his own uninhibited car-making concept.

ROBO-01: AB noodles of a catfish

The concept car is often the carrier of the brand concept, and its most conventional operation is to fly in the sky.

Therefore, ROBO-01 in itself, in fact, there is nothing worth investigating, no B-pillar, carriage doors, electronic exterior mirrors, lights, steering wheels, etc. These so-called highlights are actually commonplace in concept cars.

In the final analysis, the reason why ROBO-01 can explode is from the flag that human creatures like to see.

Because the set degree to achieve the concept car and the subsequent mass production car has a 90% similarity, but also added two premises, the first mass production car more than 200,000 yuan level, the mass production model limited model can be unveiled this fall.

This is obviously quite a dramatic tension structure.

*A-side of catfish:

Based on these terms, we still have to dissect the sparkling concept car of the set degree.

The first is the 10% to be cut, and the ROBO-01 actually has a lot of foreseeable controversy in mass production, such as unreasonable and costly front row butterfly doors, no B-pillar under regulatory risks, streaming mirrors, retractable steering wheels, and the telescopic lidar scheme that Li wants to mention.

But of course, no matter how the concept car is cut, it is still a concept car of bull X bombardment, if you can really retain the overall situation, the ROBO-01 mass production version should indeed not be underestimated, so let's look at some of the highlights that can be theoretically available.

First of all, in terms of intelligence, 3D boundless integrated high-definition large screen, which does have a high degree of mass production rationality, as an important channel for Baidu's scientific and technological strength display, the design concept of the screen is very risky to cut a shot and move the whole body, and the difficulty of carrying is not extremely high, so the possibility of retention is still quite large.

On intelligent driving and cockpit hardware, 31 sensors, including 2 lidar, two NVIDIA Orin X chips, and Qualcomm's fourth-generation Snapdragon 8295, can only be said to be quite luxurious.

There are also improved pulling demands such as speed tail, U-shaped steering wheel, light language, rear carriage doors, etc., although they are not the first of the set, and the models on the market equipped with these features include Porsche Taycan, Tesla Model S, Gaohe HiPhi X, Rolls-Royce Cullinan and so on.

But friends, the set degree is more than 200,000, and what bicycle do you want.

Although this kind of thing will also be controversial, such as Gaohe Guo Tiefu's "collection of achievements" and "jealousy" of the set degree is very disgusted, but in Tan Qing said AI's view, the scarcity of product power, or the product elements after the direct harvesting of the luxury car brand market education, are often the place where the new forces love the whole work, on the basis of not infringing patents, in fact, there is nothing to blame.

Just like Xiaopeng P7 big show scissor door, or a new brand like Gaohe, there is actually no brand power in the early market, but it can play against the first comers such as Tesla Model X in the pure electric market, largely because of the many differentiated competitiveness at the product level, just like the million-level luxury car elements such as the carriage door, Gaohe will play it to within a million.

If in a few months, the contemporary luxury car elements such as the lamp language, carriage door, and U-shaped steering wheel can be replenished in the price range of more than 200,000, then at least for the consumers of the concentration, it is not a bad thing.

The main question now is, can the set degree really be done?

*B-side of catfish:

The ideal is very full, but the words still can not be said too full, if the theory can have a high brightness set really all to give, the cost probability is unable to withstand the price of 200,000, so ROBO-01 will cut which, will it overturn after cutting? You can only look at the "fruit" after autumn.

It is not difficult to find that in the current set of flags, the biggest contradiction is actually the contradiction between the outrageous trench + intelligent gold content of the concept car and the low price of mass production and the tight time.

Looking at jokes as an instinctive need of people, the set degree flag has an important contradictory element that will collapse, which inevitably becomes the driving force for many people to pay attention to the first car of the set degree.

However, Tan Qing said that AI believes that the fact that the set degree flag does not collapse itself will not have too much impact on the set degree. For example, a company claims to build a mobile phone that only sells two thousand pieces and hangs all its peers, but finally creates a product that can only be equivalent to the iPhone, although the flag collapses, it will not stop its big sale.

Therefore, the flag of the 90% mass production concept car of Jidu, as a dispute that has both contradictions and room for space, is still a bit interesting from a marketing point of view.

On the contrary, when Talking about Ai and many industry insiders, Wenwen (pseudonym), an analyst in the automotive industry, talked about a hidden concern worthy of attention, "It is reasonable to say that the set degree is not far from the release of mass production cars, if you take advantage of this concept car conference, it is actually a very good time to warm up the core competitiveness of the product." ”

"For example, intelligent, are about to be mass-produced, in the concept of the car to do some basic UI display is certainly no problem, or some representative interactive experience, can make everyone's understanding of the car robot more intuitive and profound, or conventional disclosure of some of the vehicle mechanical parameters with selling points, but these sets in the conference and many media teachers experience the concept of the car real car, did not do."

It is also because of the lack of authenticity anchors, Wenwen is worried that many original users of Jidu will link the product concept of this conference with the actual ability of subsequent mass production vehicles, resulting in a psychological gap, which may also cause a preconceived unfavorable impression of Baidu's intelligent achievements for the first time.

But today, when the bullets are still flying, who can be sure of anything?

Instead, you may wish to think about a problem, eagerly looking forward to mass production, the market positioning of the volume, the debut is the ambition of the king explosion, and the risk of the gap is also to rush the set, what is he in a hurry?

We may be able to glimpse the clues in Baidu's needs.

Jidu Battle: The chess eye of Baidu's ten-year chess game

Baidu's three major business segments, mobile ecology, intelligent cloud, and intelligent driving.

The mobile ecology is already a mature cash cow business today, but the epidemic-caused advertising market sluggish and high-pressure Internet policies, not sexy stories, stirring up the G-spot of the capital market, just like Yang Guifei wearing a hurdle vest with four words of labor pacesetter can not seduce An Lushan and Li Taibai;

Intelligent cloud business growth rate is fast, accounting for the core business revenue is also rising, today Amazon, Ali are also doing, although the market space is not small, Baidu intelligent cloud growth rate is also very fast, but the premise of fast growth rate is that the end depends on the small base, and 0 yuan standard, one yuan standard within the roll, the distance to become a real pillar industry, there are many uncertainties.

Intelligent driving Baidu's large investment, long front, and huge future market prospects will hopefully be the key to supportIng Baidu's future valuation.

*Apollo Troika Closed Loop:

After the release of the first quarter of last year's financial report, Baidu Chairman Robin Li clarified for the first time in an internal open letter three business models of Baidu's Apollo intelligent driving business:

The first is to provide Apollo autonomous driving technology solutions for OEMs to help car companies quickly build automatic driving capabilities; The second is Baidu car manufacturing, integrating Baidu's innovation in automatic driving end-to-end, and bringing the most advanced technology to the market for the first time; The third is to share unmanned vehicles.

In addition, Baidu currently has ToG's smart transportation, which is difficult to call a business model in the early stage.

It is not difficult to find that Robotaxi and intelligent transportation are superstructures, and empowering car companies and car manufacturers is the key to Baidu's most promising results.

However, to empower car companies, Baidu is not easy to go.

Empowering car companies, Baidu is essentially a bit like L4 companies doing L2, lack of engineering capabilities, and lack of hardware foundation compared with empowering giants such as Huawei.

The problem with Robotaxi is that the battle line is too long, Waymo and Baidu have done so many years and it is difficult to land on a large scale, imagination and prospects have always existed, but R & D investment is a bottomless pit.

Therefore, car manufacturing can enable car companies to sharpen the foundation of engineering capabilities and hardware, integration capabilities, etc., and if Jidu can sell well, it will further endorse Baidu's intelligent driving technology. When the two businesses are lively, they can lay eggs along the way, take Robotaxi with them, and then revitalize Baidu's entire intelligent driving business model.

Baidu's internal emphasis on set degree is also visible to the naked eye, Robin Li OKR, the intervention of set degree is obviously greater than Geely.

It is not difficult to find that this appeal has a premise for Jidu, that is, it cannot be limited to small and beautiful, so this may also be why Jidu's first mass production car is priced in the range of more than 200,000, and sales are a very core appeal of the Jidu brand, or even one of them.

* Concept and true kung fu

In terms of capital markets, in February 2021, before the lunar Chinese New Year's Eve, Baidu's market value exceeded 100 billion market value, which was once briefly brilliant and showed signs of secondary recovery. A marketer who has served Baidu said to Tan Qing that AI said: "The article should be fast, write quickly, send it out on the same day, and don't be embarrassed if the market value falls the next day."

For the undervalued market value, Baidu's marketing level of communication is almost obsessive, telling Baidu's seriously undervalued articles, which have never been interrupted in recent years.

In tan qing said AI view: "the fall of the mobile Internet, the collapse of the C port monument, the patience of users and the capital market is exhausted, now look, Baidu's overall strength, especially the market value level, is undoubtedly undervalued, the lesson that should be made up, the debt that should be repaid, Baidu is looking for a supplement on AI"

When the shareholders are collectively optimistic, they are overvalued and have bubbles; The overall is not optimistic, there may be underestimation, there are depressions.

This is the norm of human nature.

In order to completely ignite confidence and summon the true strength of Baidu, the C-end automobile business is the finishing touch.

In terms of mobile ecology, tinkering on a broken ship is far less exciting to the crew's determination to sail than finding a new ship, and the story of traffic is more impressive in the new scene of the car.

In terms of AI business, with the signboard of the C-end business of automobiles, it is possible to better empower B-end enterprises, after all, all B-end businesses must be implemented to C-end individuals in the end.

For Baidu, the real appeal will not be how much profit Baidu brings by selling cars, and sales are the only KPIs. After the real listing, bold prediction, if the sales volume does not meet expectations, the probability of the set will choose to make profits, further, and even high subsidies to do the sales to further prove this thing, Baidu can play the closed loop.

From Baidu's basic disk point of view, sales went up, the market value can be high, the capital market is popular, in order to have more money to develop AI investment, AI research and development investment is high, Huawei, Tencent, Ali is also a lot, in order to maintain Baidu AI in the absolute first position in the country, need capital investment, late car update also need funds, enhance market value, for car manufacturing, AI research and development to provide a steady stream of bullets, to solve the worries is the key in the key.

However, the outlook still has to return to reality, after the release of the Jidu concept car, in the secondary market these two days, Baidu's stock price has not fluctuated much.

There may be two problems behind this, and these two problems may also continue until the next big move of the set degree appears.

First, when can the car robot arrive?

If you want to extract the core point of this conference, the concept can be said to be a key word.

It is true that the concept of automotive robots was proposed by Baidu, specifically, Robin Li believes that there are three major functions, free movement (unmanned driving), natural communication (high-order interaction), and self-growth (learning iteration).

Xia Yiping also said that the era of smart car 3.0 is the era of automotive robots.

It is undeniable that Baidu's judgment on the future development of automobiles is very good, but the question is how far are we from this era?

Capital is willing to buy for the story, but often not for judgment, because judgment itself is the work of capital.

Driverless, high-level interaction, and the free evolution of AI that abandons OTAs are still far away from us today, and the set degree has not yet given an intuitive demonstration of its capabilities.

The second is Volkswagen's demand for automotive robots at this stage:

Blood-soaked car robots are likely to become just what humans need.

But today's unmanned driving can not reach, can only be L2 with an infinite number of plus signs of the improved experience.

Is interaction now a pseudo-requirement? Today, the high-quality scenes of interaction, home, cars, etc. are still more promising than the real situation, even Baidu's own smallness, the current has not reached the epoch-making significance.

Not to mention freely evolving AI.

From this, the current reality is that although it is the era of automobile 2.0, it has just arrived and has not yet reached the mature stage.

Therefore, the current demand for automotive robots is still a paradox to a certain extent.

Left hand Baidu, right hand Geely, with the golden spoon born of the set of degree, ready-made technology, ready-made production capacity, is indeed better than many self-made car builders, with the car to drive faster chips.

All of the above, perhaps Jidu still can't put a bloody car robot and disk out, but he does lead the industry to take a small step in the possibility.

I have heard such a metaphor, saying that imagination as for the capital market, is like a faint light in a cave, making people motivated to follow, as to how long it takes to get out of the cave, it depends on luck.

If the car robot looks at it today, it is actually still talking about where the light is in the cave that we are going to go tomorrow, which has extremely limited significance for the present.

In today's cave, the closest light to us may be this autumn.

We also wish that the first shot of the Internet giant car manufacturing can successfully break the boundary between us and the next era of the automobile.