(Original title: Why is the "non-critical" AI in the field of physics the best opportunity today?)
Artificial intelligence has a variety of classification methods, one of which is to distinguish between actors and fault tolerance. Artificial intelligence is divided into the physical domain and the virtual domain, and the two have their own key and non-critical differences.
Although this classification model has a very ambiguous boundary, it can be confused with the application scenario to create an ambiguity in the underlying technology. However, it is relatively valuable for commercial applications.
It should be noted that although a lot of public opinion focuses on key areas AI, such as driverless, artificial intelligence medical, and financial AI. However, the commercialization of AI in key areas still has a long way to go. Non-critical areas are obviously ways to enter the market faster to achieve commercial transformation.
Technologies such as information flow recommendation, intelligent voice input, and face recognition are examples of non-critical areas where AI is rapidly entering the application phase. However, non-critical AI in the physical field is clearly not as hot as it is in the virtual world. The attention received is far from the same.
In particular, many so-called "smart hardware" on the market are called "smart" by placing the interactive system in hardware. These pseudo-AIs in fact destroy the basic user's basic knowledge of AI+ physics hardware.
But think about it in reverse, these bluffing smart hardware also set up a blank window for the already crowded AI market. At present, small companies and entrepreneurial teams want to enter the AI ​​field, which may be one of the best opportunities.
What is non-critical AI in the field of physics?
First of all, we can rethink what exactly is the non-critical AI in the field of physics. One well-known example is that driverless cars are AI-critical applications, while sweeping robots are non-critical AI. Because driverlessness can never go wrong, even a one-hundred-thousandth error rate is a big deal. The sweeping robot does not matter, to maintain a certain error rate is actually quite adorable.
This statement is generally good, but it should be noted that not all sweeping robots are AI, just as some scholars have proposed that drones can go to AI. Only using the AI's core algorithms to provide logic support for learning, processing, and judgment for the machine or hardware can be regarded as a true physical AI product.
It should be 14 years later, smart homes and smart wearable devices have been buoyed by the market. However, this intelligence is not smart. Using mobile phones to control air conditioners and importing data collected by smart wristbands into simple algorithms can only be regarded as the extension of mobile computing devices. It does not change the basic characteristics of its tools and autonomous intelligence.
Therefore, it is by no means possible to connect WIFI, and the interactive option is AI product.
So how do you define the basic features of AI+ hardware? The first thing to ensure is that the hardware itself has the ability to learn from the data and output independent results. For example, the robot can determine the position of the object based on visual data, thereby outputting an independent crawling scheme and operation method.
The second is to ensure that there is sufficient coordination between the AI ​​algorithm and the hardware, and the mechanical part can reflect the derivation results given by the AI. Otherwise the machine can only think but can't do it, and it becomes a computing device.
In light of these points, it is not difficult to find out that there are not many AI applications in the entire physical field. However, the application scenario of this market is far more extensive than that of key applications.
The back of the robot is more than just the Internet of Things: the real opportunity is to make life more
Since human cognition habits and works of art have been repeatedly brainwashed for a long time, and when artificial intelligence is mentioned again, everyone involuntarily thinks of robots—and this kind of association is not friendly most of the time.
In any case, after a long period of independent development in the field of robotics and artificial intelligence, although artificial intelligence has not yet helped robots become terminators, robotics has provided a lot of imagination for AI+ physical hardware.
In the implementation of robotics technology in the field of machinery, the Internet of Things is also rapidly developing, providing the basis for hardware devices to perform operations, transmissions, and multiple interactions. Under the common support of these two technologies, the AI ​​application in the physical field has a very good evolutionary basis.
We can try to give a few examples to see what evolutionary space AI has in physical applications.
1, home appliances: smart home made a lot of years, but has not opened the real consumer outlets. This is largely because consumers do not know what these "smart" can do. Appliances that incorporate chips, systems, and network ports are still unable to provide essential functionality changes, which has led to the promotion of intelligent propositions. This is also the problem that the algorithm and the hardware cannot coordinate with each other.
For example, most people don’t need to use a mobile phone to operate the washing machine or watch a match on the washing machine, but they need the washing machine to accurately wash the laundry corners, judging the washing time and the detergent types and mathematical principles—machine vision technology may be able to do To these.
2. Toys: AI+ toys seem to have been relatively awkward topics, but in fact, the judgment and perception of AI is quite extensive for toys or game devices. Whether it is the collection and processing of children's growth toys for children's accompanying toys, the ability of competitive toys to adapt to their owners and the growth of individual styles, and even the adaptation of the VR all-in-one to the human eye through AI, it is a “small technology†with extensive potential.
3, security equipment: AI to solve security topics may also be many, such as smart locks, AI suitcases and so on. Depth identification technology can better determine the identity of the owner, and even operate the security device through voice, biometrics and other methods. This increases security and also avoids embarrassment in many special situations.
4, sports equipment and health: According to Parkinson's patient's hand shake the design of intelligent anti-shake spoon was once popular. Although the technology is immature, its real application value is not great. But the logic of collecting patient or athlete data based on hardware to provide a personalized solution is proven.
5. Fieldwork: At present, more AI applications in the field of drones are in the field of image capture. But in fact, whether it is underwater underwater probes, field probe vehicles or drones, you can use AI to provide more judgment and processing capabilities. The recent drastic drone impact on civil aviation incidents is actually a small problem that can be handled with AI technology.
In summary, the real value of AI applications in the field of physics is not to allow you to use mobile phones to operate air conditioners, such as “provide more choices,†or to “give more functions†to buy eggs in refrigerators. The real opportunity lies in using AI-driven hardware to pass data processing and algorithmic judgments to accomplish what humans did not think of, or think of that they couldn't.
Just like smartphones provide services that traditional mobile phones such as payment, social networking, entertainment, and O2O cannot provide. The same logic, AI hardware should not be to save what life time and cost of living, but should make life more.
Under the monopoly of big companies, the field of physics drips out a few AI business opportunities
There is a very cruel topic that many entrepreneurs and investors are unwilling to face. That is AI is absolutely a big company game, whether it is at home or in the world.
We must know that the requirements of AI's entrepreneurial requirements for data volume and core technologies are among the highest in the Internet technology revolution. And under the condition that the world is regarded as determining the future, AI's core hardware, algorithms, talent, and even the power of discourse must be quickly divided up. AI applications in key areas, and user-oriented AI applications such as search, information, and translation are battlefields for big companies.
On these battlefields, there is no room for entrepreneurs to breathe. So non-critical applications in the physical field are likely to become an effective "back road." For small companies and entrepreneurs interested in AI, this area has at least the following advantages:
1. Vertical scenes are numerous and highly restrictive: As a philosophical thinking evolves into a technical solution, AI may have a lot of fields. Especially in the physical domain that does not rely on platforms and traffic, the application scenario is almost limitless. This gives vertical entrepreneurial opportunities, at least not quickly facing the big companies.
2. Reduced core technology requirements due to fault tolerance: AI's fault tolerance rate is a pyramid-type game. Just like unmanned driving technology, large companies and car companies can go to gambling in exchange for a few percent of the fault-tolerance rate for ten years. Decline, but entrepreneurs obviously can not. Therefore, in the area where the fault-tolerance rate is relatively high, even small companies that can make mistakes in the field will be more likely to have room for survival and admission.
3, the speed of opening the market may be fast: AI is a new mode of thinking, after clever use it can change a lot of hardware or machinery to give people common sense. This makes it possible for products to shake the market by quickly changing consumer perceptions.
4. Withdrawal channel is relatively benign: Nowadays and in the next AI business world, the completion of the acquisition of venture projects by large companies is bound to be the main theme. This provides a lot of launch opportunities for entrepreneurs in the field of bypass physics. In fact, this is important for most entrepreneurs.
5. The open source platform is lowering the threshold: Although core technologies and production capabilities are at the top of the pyramid, open source platforms such as deep learning are emerging. What big companies need is the ecology of AI entrepreneurs, and the physics field is obviously that Internet giants who master algorithms are not easy to get involved.
6, combined with manufacturing can strengthen barriers: immediately above this, AI's core players are Internet companies and ICT companies, their weakness lies in the alienation of the manufacturing industry. The combination of manufacturing and AI is also an effective way for entrepreneurs to strengthen their own industrial barriers.
Having said so much, the core is actually very simple: AI is not a beautiful world, but a cruel jungle battle. To divide the future into a cup, we must drill through the gaps to the future.
Cases and Inspiration
Finally, I want to mention a very famous robot startup company, Rethink Robotics of the United States. The company's Baxter collaborative robotic solution is dedicated to improving industrial productivity and is well known in the robotics field.
But it is interesting that Baxter’s real market advancement has not been good for several years. A lot of voices on social media have shown that this guy with a stunned look can really do very little. The factory did not know what to buy. Instead, it has a good reputation among scientific research institutions, colleges, and entrepreneurs, and many are willing to buy back at high prices for pure research.
This is because Baxter has established an open source hardware platform. Other developers can easily edit their own robot systems and human-machine interactions based on other platforms to solve various problems in vertical scenes.
This logic may also hold in the next AI+ hardware area. Based on the open source platform, the research and development of more vertical products, combined with their unique needs and hardware production capabilities, may be a feasible solution to open the AI ​​entrepreneurship channel.
The picture of AI's complete subversion of human life may be only a matter of time. But before that scene came, gradually allowing more people to adapt to AI and feel the existence of AI is today's "epic task."
It may be like chasing a girl to welcome the arrival of AI. Perhaps it is not possible to buy luxury villas such as "driving without a driver" and "quantum computing" today. Of course, it is somewhat strange to talk about them. But writing a love letter or buying a small gift might be okay—the value of non-critical AI in the physical field is just here.
燑br>
Most Popular Vape Puff XXL 1600 Puffs Bar Disposable Vapes 1600 Puffs Vape Cartridge Packaging
Features of disposable vapes:
1, The disposable vapes give you 1600 Puffs per bar
2, 5ml pre-filled e liquid, 2% Nic strength, 800mAh pre-charged battery
3, Draw activated mechanism, simple and happy vaping exprience
4, Life Expectancy is 24 months.
5, This disposable device requires no maintenance, charging or refilling.
Flavors:
1. Strawberry Pitaya
2. Ice Peach
3. Ice Grapes
4. Mango Smoothie
5. Nutty Tobacco
6. Pink Lemon
7. Apple Peach
8. Blueberry Raspberry
9. Pog Orange Guava
10. Fresh Milk
Disposable Vape Puff,Electronic Vape,Disposable Vapes 6%,Vape Electronic Cigarettes,Electronic Cigarettes Vape
Shenzhen Axiswell Technology Co., Ltd , https://www.vapingera.com