Recently, many organizations will put on a very professional posture and tell you how to simply learn ChatGPT to earn 1 million or 10 million.
You can't make money because you don't work hard.
They'll make you pay to explain what ChatGPT is within minutes of chatting with you, telling you that you need ChatGPT's work.
In their eyes, ChatGPT became a god-like money-making machine.
But no one ever asked you to search Google, and OpenAI, the company that developed ChatGPT, is still not profitable.
OpenAI has been acquired by Microsoft in recent years, and Microsoft CEO Bill Gates indicated the promise of artificial intelligence, but did not immediately indicate what kind of money-making model exists for ChatGPT.
Perhaps those who rely on ChatGPT to earn course fees tell you that the state refuses to teach its money-making skills to ordinary people because of the stakes.
But......
Aren't they similar to pyramid schemes?
Even they themselves do not understand what ChatGPT is, that is, they start educating you how to make money through ChatGPT, if you can't make money, it means that you have the wrong method and do not work hard.
It sounds like the logic is right, but does the internal logic in it have anything to do with it?
The biggest correlation is to make you pretend to work hard, right?
Paid the money, solved a few days of anxiety, did meaningless things, wasted time, and then got worse and worse.
You don't need to understand the technical, commercial, and industrial chain logic of AI to start making money.
Many people who don't understand AI are using AI as a marketing tool and you as prey.
They know better than you that they don't understand AI.
The promise of AI is there.
But not as the media portrays.
At present, a very funny scene in society is that a group of people who have never been in contact with artificial intelligence sit together, make a big fuss, say various professional words, and one by one like people who are about to earn 10 million.
Trouble these people to turn left to buy lottery tickets.
Or buy at the highest point of Bitcoin and experience what it means to be real.
On the contrary, engineers, architects, and programmers in the field of artificial intelligence sit together, talking about sensors, algorithms, and application scenarios, looking at each other, although everyone did not say anything, but in their hearts they would say: "How to make money with this thing?" ”
People in specialized fields and people who don't understand AI at all, the AI they are talking about is actually two things.
AI in 2023 is in the early stages of technology and in the early stages of commercialization.
AI will not become human, and the IQ of AI in the future may be hundreds of thousands of times higher than that of humans, but the thinking model of AI and humans are still different.
AI can indeed replace teachers, doctors, and programmers, but not as everyone thinks.
It is to replace some meaningless simple and repetitive work procedures.
For example, teachers use the same way to educate students with similar problems in the classroom, as long as the student's growth environment, problems, behavior patterns, contradictions and other key factors are entered, the robot teacher with the AI system can interact with the student through voice and propose solutions to help the student solve the current dilemma.
Then the teachers themselves can free up their time to do more important things.
Without the participation of hardware devices, do you think such classroom AI can be achieved?
Can you just take a ChatGPT and issue a text, can you solve the problem?
You know better than anyone that as a doctor, as a lawyer, as an aircraft repairman, you face intractable problems that cannot be solved by an AI conversational problem.
ChatGPT gives a picture of the future, pointing to the direction of future technology and business, as well as potential possibilities.
Instead of being able to commercialize it now.
Those artificial intelligence tutors, even if they don't use the topic of AI, can be replaced by quantum mechanics, which can also make you pay for the class.
You think about relying on others to solve your own life difficulties, and people want to empty your wallet.
Mentors who really understand AI, have long stopped doing AI consulting, real money in hand, last fall to buy NVIDIA US stocks, on October 14 last year, NVIDIA US shares 112.27 US dollars, and then all the way up, on May 30, 2023, NVIDIA US shares 401.11 US dollars.
If they understand AI, it's not difficult to understand NVIDIA, right?
But they may not even know what Nvidia does.
Founded in 1993, CEO Jensen Huang is a Chinese and the only Chinese among executives, he speaks standard American English, knows Hokkien, can speak a little Mandarin Chinese, and integrates well into the American community.
The company's main semiconductor business, this year in the direction of artificial intelligence, recognized by capital, stock prices rose, recently reached the first trillion dollar market value, and Apple, Google, Amazon and other head technology companies reached a similar level.
The direction of AI is not simply chatting tools.
Instead, it is combined with hardware to go deep into all walks of life in the process of continuous evolution of the model and propose solutions to real-world problems.
The semiconductors launched by NVIDIA are suitable for promoting the development of artificial intelligence, not only algorithms, no storage, no carriers, and no efficiency.
The future of artificial intelligence pays attention to the simultaneous existence of 4 elements:
1. The ability to mine and analyze data.
Compatible with online and offline mining, which requires the ability to collect data on the global Internet, as well as the ability of hardware devices to collect offline.
For example, in the classroom, hardware devices collect children's expressions, health and other information for a long time, and analyze them, through artificial intelligence to predict which child has school violence, which child has psychological problems, which child's native family has problems, after issuing a complete report, AI continues to give a plan for how it will intervene.
Similarly, in law firms, insurance institutions, banks, ocean freighters, and cross-border e-commerce, artificial intelligence realistically collects all important information online and offline.
Without hardware equipment, how to collect information such as humidity, wind direction, and ocean currents on the ocean, without this information, how to make timely safety risk predictions.
2. The ability to interact with the real world and turn the output information into action.
For example, robotic arms, sweeping robots, drones, need the participation of artificial intelligence, the output information into action, so that industrial robots in the factory to operate, the links are more complicated, how to make the joints of the robotic arm more flexible, this is a special headache for engineers.
3. The ability to integrate the supply chain.
From AI algorithms, to autonomous driving, to new energy vehicle manufacturers, some companies do voice recognition, some work to do facial recognition, and different parts are integrated together to become a complete product for consumers to use.
In a supply chain, there may be dozens of companies, or even hundreds of companies participating in the cooperation, without which the product is incomplete.
4. Find monetized application scenarios
This is the most troublesome problem for various artificial intelligence manufacturers, relying on semantic analysis and chat software alone, there is no way to monetize, these are not real application scenarios.
SenseTime and iFLYTEK now lack application scenarios that can be replicated on a large scale, and the algorithm technology provided by SenseTime is very high, but the smart city products needed by government departments are not provided by SenseTime.
SenseTime is a leading AI manufacturer in China, and even this company has not found a particularly large-scale application scenario, let alone the same type of manufacturer.
Application scenarios that need to solve problems are in demand in all walks of life, but no AI company provides them on a large scale.
If you figure this out, you know that the AI mentors who tell you that they don't have ChatGPT works are purely scammers.
If you cultivate for a long time in your own professional field and do it well, you will naturally be able to earn money.
It's not that you leave the house halfway, you don't understand this industry, you start to invest money, and before you understand how you are deceived, the money is gone.
Nvidia CEO Jensen Huang recently said to the media that we either try to be predators or try very hard to become prey.
At least in the vast majority of the crowd eager to make money, most of them have become prey.
From the MLM in the late 90s, to the direct sales later, to the quantum mechanics training course, it also includes some milk tea brands to join.
They've always been there, they've never made money.
Prey can't make money because they work very hard to give their wallets to others, and they try very hard to imitate all the behavior of rich people, except for not doing things in their own industry.
I used to think that Dong Shi Xiaoying was a joke.
Today, many years later, I found out that Dong Shi Xiaoying is a reality.
Rich people can become rich, relying on the core logic of making money and the long-term patience of deep cultivation of an industry, instead of learning this asset today tens of millions of people go out to take taxis, tomorrow learn that asset billion people to say witticisms.
Armstrong landed on the moon and said, "Today is a small step for me, but a big step for mankind."
If you say this on the bus, the people behind you will have to push you and scold by the way: "Hey! Are you walking or not? ”
There are many areas where AI can make money, but to open up this supply chain, any of them is very difficult.
Taking the more popular smart city business that has been speculated in the past two years, for AI companies, it is necessary to access the webcam, and then identify the person in the lens, determine whether the person is fighting, whether he is smoking, whether he is running a red light, whether he is not wearing a hard hat and a series of specific scenarios.
Analyze the objective results, improve the precision and accuracy, and then repeatedly negotiate with the customer, and then push the results.
Do you think that such a scene is particularly boring, yes, the things that can really make money are very boring and boring in themselves.
What can really create profits often seems to be very boring, which is often something that most people who want to make money do not have the patience to understand.
When making this artificial intelligence product, you need to have pedestrian detection, face key point detection, face recognition, license plate detection, and license plate recognition models.
It is also necessary to deploy various deep learning models, and instead of using the original model to run data, because this speed is very slow, in order to enhance efficiency, hardware devices and AI cooperate to continuously collect and analyze more data.
To solve these problems, it is necessary to find engineers who understand the underlying architecture, but also understand the upper-layer algorithms, and also write the heterogeneous code of CUDA by hand.
We only talk about the direction of smart city, and we must implement network streaming media, video codec, front-end, UI, back-end development, alchemy, algorithm deployment, heterogeneous development, operation and maintenance, etc.
If you also want to do the mobile phone end, you must also design for Android, iOS systems, etc., do automatic driving, but also involve ROS, industrial Internet AI to involve RTOS/Qt.
These require the use of different technology stacks, such a complex team, and involve payroll management, R&D management, cost management, and operation and maintenance management.
Those who do not understand AI think that AI is a chatbot, and think that chat can solve all problems.
Psychological counseling is chat, but do you think AI can replace counselors with AI chat in a short period of time?
Think about it, you yourself know it's impossible.
Teacher Huang, who has recently become very popular on Douyin, performed "digging and digging" in the video.
Based on the AI of the kindergarten scene, that involves more things.
AI is used for monitoring at the entrance of the kindergarten.
AI will be used in canteen cooking robots and food safety detection.
AI is needed to observe children's sleep quality, expressions, voices, and body movements.
The analysis and observation of children's native families requires the use of AI.
Kindergarten teachers need AI to join in the evaluation of children's education methods to help teachers improve the way they teach children.
Just a kindergarten is so complicated, let alone elementary, secondary, university, adult education, interest education, private and kilometer schools, dance, yoga teaching, etc.
This is a huge AI industry chain.
This is just an education system.
In larger cities and villages, it includes all walks of life, and a larger super AI industry chain containing different subsystems.
Speaking of which, do you still think you know AI?
We better be in awe of new technologies.
The development trend of technology is unknown, the business model is unknown, we need to keep trying, constantly thinking, and constantly discovering ways to make money in the field of AI, rather than directly believing the media's "someone has already made money with ChatGPT" to comfort themselves.
Even SenseTime has no way to achieve profitability in the short term, let alone a layman who does not understand AI.
In 2018~2022, SenseTime's operating income was 1.853 billion yuan, 3.027 billion yuan, 3.451 billion yuan, 4.700 billion yuan and 3.809 billion yuan, respectively, and the year-on-year growth rate from 2019 to 2022 was 63.30%, 14.02%, 36.20% and -18.97%, respectively, it can be seen that before the negative growth, SenseTime's revenue growth has begun to decline.
SenseTime's financing ability is very strong, and the company's technical level is very powerful, but their technology cannot find suitable application scenarios, and it is difficult to replicate and promote on a large scale.
The company's stock price has plummeted since the beginning of listing, as shown in the following figure:
Finally, let's look at how the AI risk control system rates SenseTime:
Technology and business trends leave some traces of looking at the future, but we need to focus on them and discover them slowly.
It is believed that AI will lead technological innovation, but the business model, application scenarios, and technological innovation in it need to be explored repeatedly.