Source: Intelligent Investor
In a recent investment strategy meeting by ARK Invest, renowned venture capitalist and founder of startup incubator FoundersSpace, Steve Hoffman, was invited to share some "brain-teasing" insights on the big picture of AI.
Hoffman is known as the "Godfather of Silicon Valley Venture Capital" not only for his successful practices in startup incubation but also for his influential works such as "Let the Elephants Fly" and "Crossing the Winter," which have impacted many entrepreneurs.
He is a seasoned startup mentor and technology trend observer, as well as an expert well-versed in the ways of venture capital.
During this exchange, Hoffman showcased many new developments, especially the exciting latest trends in Silicon Valley, which will have profound impacts on current businesses and the economy.
For example, he mentioned:
"(In Silicon Valley) there is almost no middle ground. Either you can't raise a penny, or the funds you get are astonishingly large. Nowadays, all the money and all the venture capital institutions are pouring into a few hot startup companies like crazy."
"If you observe the investment portfolios of Silicon Valley venture capital firms, you will find that 2025 will be a year of success or failure for enterprise-level AI."
The two key strategies of open-source and distillation have completely changed the competitive landscape of the field of Artificial Intelligence.
An increasing number of companies are beginning to innovate around small models, which is a huge growth area and also brings many new investment opportunities.
The AI agents being developed are the first systems in human history that can truly perform complex tasks and replace human labor.
Robotics plus AI is the future direction of human society. We are approaching an unprecedented critical point.
Smart investors do not place all their bets on one technology; they diversify their investments and wager on different possibilities.
Hoffman mentioned that a large number of new AI technologies and products are emerging, emphasizing that the reality in the wave of AI is that what we consider 'certainty' today may be completely different tomorrow.
Although these visible innovations are accelerating, in the next 10 or 20 years, there will almost no longer be traditional jobs, as robots will be able to perform all tasks at lower costs.
However, Hoffman is not conveying pessimism; he believes that AI's reasoning methods are fundamentally different from humans, so 'AI will not become true humans,' but we must confront the challenges, especially understanding the future trends of wealth distribution.
"If you are managing your own wealth, you need to know the direction of future development. All of this will happen in our lifetime, and those who control Robotics and AI will control most of the wealth in the world, which is an undisputed fact."
Hoffman also provided a list of key factors for identifying the investment value of numerous AI companies.
Smart investors have organized and shared the most important content from this engaging speech. Enjoy it~
1. We are at the initial stage of the AI wave.
Today's theme focuses on exploring the key field of Artificial Intelligence and Robotics technology, carefully understanding the trends in investment and technology both in Silicon Valley and globally.
Currently, the field of Artificial Intelligence is experiencing rapid explosive growth, with thousands of new companies emerging daily. In fact, most of these new companies struggle to survive beyond five years.
Market competition is extremely fierce, with many companies lacking truly valuable core technologies. Many startups rely solely on existing platforms such as ChatGPT, Google Gemini, DeepSeek, etc., using their APIs without independently developing AI technology or having exclusive data.
These companies may seem strong at present, but they can easily be replaced.
Therefore, when considering investing in such companies, investors must possess profound professional knowledge and sharp judgment.
It cannot be concluded that a company has inherent value in the future just because it is involved in the field of Artificial Intelligence and its business seems impressively new.
Many people ask whether there is currently a bubble phenomenon.
We are familiar with the '.com bubble' of the Internet era, but the situation this time is different because the Artificial Intelligence field is indeed creating value.
If a company does not adopt Artificial Intelligence technology, it will find it difficult to stand in the market in the future. Artificial Intelligence, like the Internet, is a foundational technology.
It can make business operations smarter, automate processes that were previously difficult to automate, enhance efficiency, and even predict future development trends; this is the power of Artificial Intelligence.
However, at the same time, some companies are severely overvalued, and their market value far exceeds the level that can actually be achieved in the future, making investment in such companies an unwise choice.
Nevertheless, from an overall perspective, Artificial Intelligence will continue to transform the business world, and within the next five years, it is expected that numerous new unicorn companies will emerge.
Currently, we are at the starting stage of the artificial intelligence wave, and this revolution is far from over; the road ahead remains long.
2. Current venture capital trends in Silicon Valley.
The current situation in Silicon Valley is actually different from what everyone imagines.
Some may believe that there is still a large influx of funds into the artificial intelligence field in Silicon Valley, and investment enthusiasm remains high. However, in fact, compared to 2021, the capital size of venture capital funds has shrunk by 60%.
For new funds, fundraising difficulties have significantly increased, with the fundraising amount for first-time funds dropping by 72% year-on-year. Currently, only more mature and well-established funds can still successfully raise funds.
Many unicorn companies still appear glamorous on the surface, but in reality, most of their valuations have declined.
The market is undergoing changes. The main reasons are: firstly, the exit cycle for companies is constantly lengthening; secondly, many companies are at high valuation levels. High interest rates have also had a significant impact on the venture capital market.
Venture capital institutions hold a large investment portfolio but find it difficult to achieve exits. Their limited partners (LPs) have also been waiting for more than ten years, eagerly hoping to get their funds back.
Some venture capital firms are attempting new initiatives, starting to return part of the funds to LPs while selling off some of their shares to private equity firms.
In addition, a new trend is emerging in Silicon Valley where venture capital institutions are starting to establish their own startup companies, acting as 'incubators', which is not the traditional model of Silicon Valley.
The reason is that the valuations of high-quality companies are already very high, and with intense competition, venture capital firms find it difficult to invest sufficient funds, resulting in unsatisfactory investment returns.
Venture capital institutions are therefore changing their strategy; rather than competing among numerous venture capitals, it is better to create AI companies themselves and take control.
If one observes the investment portfolios of Silicon Valley venture capital firms, it will be evident that 2025 will become the 'year of success or failure' for enterprise-level AI.
All venture capital institutions and the entire market are closely watching these AI startups; can they truly achieve profitability? Can they really obtain the enormous profits expected by the market?
Currently, the profitability of AI directed towards consumer markets is limited and has not yet yielded significant returns.
As a result, current investors in Silicon Valley are betting on enterprise-level AI, which refers to artificial intelligence products and services provided to Fortune 500 companies.
But the question is whether these enterprise-level AI companies can be profitable as expected.
There is currently no conclusion, which is precisely the focus of venture Institutions right now.
3. A handful of companies have attracted a large amount of funding.
At the same time, we also see another trend: now more and more small teams are only receiving limited funding, but with efficient use of Artificial Intelligence, their employees can maintain high productivity.
Small entrepreneurial teams with less funding can actually accomplish a lot of work. Nowadays, some investors specifically invest in these early-stage startups.
Even if these companies eventually exit at a smaller scale, such as only 0.1 billion dollars instead of a 1 billion dollar unicorn valuation, they can still achieve a good return.
In Silicon Valley, I often tell the founders of startups that you are either in a situation where funding is nearly impossible, or you are so hot that all investors are scrambling to give you money.
There is almost no middle ground here. You either can't raise a penny, or you receive an astonishing amount of funding.
Currently, all funds and all venture capital institutions are pouring into a few hot startups.
For example, one typical case is Dev Agents.
Dev Agents has just completed a seed round of financing, raising as much as 56 million USD. This company has not really proven itself yet, just getting started, but its valuation has already reached 0.5 billion USD!
One reason is that its founder was the CTO of Stripe.
Stripe is one of the largest payment platforms in the world, performing extremely well. This former CTO is well-known in the industry, has strong connections, and possesses grand vision.
He plans to develop a brand new cloud-based operating system using artificial intelligence, aiming to create an operating system that can compete with Apple, Google Android, and Microsoft Windows, but this operating system will be entirely cloud-based.
All operating systems we currently use are device-based, running on specific hardware.
So, in Silicon Valley, if you have a reputation, a strong background, and can paint a sufficiently grand vision, you can still secure massive funding.
However, this also reflects the characteristics of the current market—a few companies attract a significant amount of funding, while most other startups face funding shortages.
4. Two breakthroughs of DeepSeek.
DeepSeek is not only a phenomenal presence in China, but also globally.
In fact, many Americans are using DeepSeek. Its application has been installed on many people's phones, and many companies are also utilizing its open-source model. The impact is tremendous.
A year ago, everyone was still asking: How will China catch up with the United States? The leading advantage of artificial intelligence in Silicon Valley is so great, can China catch up?
At that time, I answered in my speech: "I cannot tell you when China will catch up with Silicon Valley, but I can tell you that the fundamental transformation in China's artificial intelligence development model will completely change the competitive landscape."
In fact, this transformation could happen overnight, and that is exactly what we are witnessing.
The first breakthrough of DeepSeek lies in not building a massive large language model (LLM) from scratch. Training such a model is extremely costly, involving billions of dollars in investment, including data collection, processing data, training the model, and so on.
On the contrary, they adopted the approach proposed by OpenAI.
OpenAI has publicly stated that you do not need to train a large language model yourself; you can create a smaller, more efficient model through a process called 'distillation.'
'Distillation' is a very important term because it fundamentally changes the landscape of artificial intelligence.
The core principle of distillation is that a small AI asks questions of a large AI, learning from interactions with the large model to gain abilities close to that of the large model. In other words, the large AI is trained at a great expense, while the training cost for the small AI is much lower.
DeepSeek is the first company to realize that high-performance models can be trained at a very low cost using this method. They are doing just that.
DeepSeek's second major breakthrough is that they have open-sourced their model. This is crucial.
Previously, Meta had begun to open up part of the source code of its AI models and gradually narrowed the gap with ChatGPT. However, DeepSeek's strategy is to directly make the trained model available for everyone to use.
Therefore, the combination of open-sourcing and distillation has fundamentally changed the competitive landscape in the field of artificial intelligence.
We can now see that Chinese companies are able to compete with Silicon Valley in the field of artificial intelligence while investing far less than American tech giants.
In the future, we will see more 'DeepSeek' emerge. You might be using DeepSeek now, but in the future, any company can replicate this model.
In the future, we will see a large number of startups utilizing this method to create very powerful artificial intelligence systems without significant financial backing.
This will completely change the industry landscape. Smaller companies now also have the opportunity to challenge industry giants.
It is worth mentioning that the AI technologies we are discussing are all text-based artificial intelligence, without involving audio or video. The technologies in these fields are completely different.
Especially video processing, which requires much higher computational capabilities, demanding extremely high computing resources, a lot of servers, and powerful processing capabilities.
So, the game is not over yet.
Platforms like ChatGPT, Gemini, etc. still hold significant advantages in the video AI field because they possess large cloud computing platforms that can host AI and provide strong computational power.
5. Six key factors for assessing the investment value of AI companies.
A key question to discuss is: how do you determine whether an AI company is worth investing in when evaluating it?
To determine whether an AI company is outstanding, the following points are key factors:
First is the team. The reason DeepSeek's team has been able to rise rapidly is that they possess top talent. They can quickly gain insights into OpenAI's practices and find their breakthroughs based on that, successfully executing them. Therefore, the team's abilities are a core indicator in judging the potential of an AI company.
Second is proprietary data. This is data unique to the startup or business, which others cannot easily access. Furthermore, this data must truly create value for customers.
As an investor, you need to determine: does this AI company possess data that cannot be replicated? Can this data be transformed into actual business value that makes customers willing to pay?
If an AI company can use proprietary data to build competitive barriers and create value that customers are willing to pay for, it may be a company worth investing in.
Third is the innovative business model. If this AI company is exploring a completely new business model that others have yet to venture into, it has the opportunity to seize a market advantage.
Fourth is patent technology. Did they apply for patents that prevent competitors from easily copying their products?
If the company possesses exclusive technology that can build a strong moat for the business, then its investment value will be higher.
Fifth is the network effect. This means that the more users a system has, the greater its value.
If a platform similar to the market can be built that attracts more buyers and sellers, then its competitiveness will continuously strengthen; alternatively, if an AI ecosystem can be established that attracts partners to join, its market position will become increasingly stable; a social network can also be created to connect people together.
Sixth is the brand. Why do most people use ChatGPT? Because they have heard of it, it is a brand. DeepSeek was able to rise rapidly because it was the first small company to challenge industry giants, quickly establishing a strong brand influence, and this brand itself holds great value.
Therefore, when analyzing a startup, the more of these factors they possess, the greater the investment potential.
6. Investment opportunities in small models and vertical fields.
Another very interesting area of innovation is small language models.
We have all heard of those large, expensive language models that require a massive amount of data for training, almost covering the entire Internet.
Most businesses in the future actually do not need general AI; they need AI that can perform specific tasks for them, which is where small language models become extremely valuable.
Training data for small models in specific fields can deliver excellent results, and in certain narrow professional areas, they even outperform large models.
If you are a manufacturing company that needs AI to optimize production lines, you do not need data from around the world; you only need data relevant to your manufacturing business, and these small models can run at a very low cost.
Another area where small models are very valuable is internal deployment within companies.
For example, in China, every bank must ensure that data is autonomous and controllable; they must run AI locally, so they will naturally choose small models to operate in their own data centers.
Because of this, more and more companies are starting to innovate around small models, which is a huge growth area and has brought many new investment opportunities.
Another area of focus for venture capital at the moment is vertical AI.
Vertical AI refers to AI that is targeted at specific industries. It is not ChatGPT or DeepSeek, but AI that solves industry-specific problems.
For example, the hotel industry needs AI to optimize hotel management, mining needs AI to improve exploration efficiency, agriculture needs AI to manage crops and pests, and the transportation industry needs AI to optimize logistics, autonomous driving, and scheduling systems...
In all these industries, what they need is specialized AI, not a general-purpose large model.
If you can be the first to apply AI in a certain industry and gain a dominant position, then you will be in a very advantageous position in the future.
7. AI agents are the most important trend.
Currently, the most important trend in the field of AI is AI agents.
AI agents and chatbots are two different concepts.
A chatbot, for instance, when you converse with DeepSeek or ChatGPT, you usually ask it a question, and it gives you an answer. It can help you complete certain tasks, like writing an article or generating a video, but it handles single tasks.
AI agents have surpassed this point. The role of an agent is that it can perform more complex tasks rather than just a simple single-step task. You can ask it to complete a multi-step, complex task, and it will figure out how to accomplish it on its own.
For example, suppose you tell the AI agent, "I want to travel to Silicon Valley, help me plan my trip." This is not a single task, but a complex task that involves multiple steps.
It needs to book flights for you, pay for the tickets, check your schedule, book hotels, arrange your meetings in Silicon Valley, and coordinate different affairs.
Currently, every major AI company is developing AI agents, and every large enterprise is building its own AI agent system.
Why are AI agents so powerful? Because they do not just help us complete tasks; they can even replace entire job positions. Whether you are a marketer, salesperson, or procurement specialist, AI agents will work alongside you and take over a significant portion of your work.
The AI agents we are developing are the first agent systems in human history that can truly perform complex tasks and replace human labor.
The hottest trend in Silicon Valley right now is not only AI agents but also groups of AI agents. This does not refer to a single AI agent, but rather a whole group of AI agents working collaboratively.
For example, suppose you work in the marketing department and you want to launch a new marketing campaign for a client. Then, you are not just interacting with one AI agent; you might be working with five, ten, or even twenty agents.
For example, Emma, a startup located in Silicon Valley, is developing a "universal AI employee." This AI employee is actually composed of multiple AI agents, each responsible for different work areas.
These AI agents can write proposals, communicate with clients, analyze data, and automate workflows. Their core idea is to allow you to delegate many tasks that were previously done by humans to AI.
Will these AI agents replace everyone's jobs? The answer is no, not happening tomorrow.
However, if you extend the timeline and look at the future in five, ten, or fifteen years, the answer is clear—most tasks that humans can do will eventually be replaced by AI.
It's just a matter of time. The trend has already formed.
Let me provide a few more examples of AI agents.
A startup named CoSign has developed an AI agent called Genie. Genie's role is that of an AI engineer, and it can write code directly.
Of course, all these AI agents sound impressive, but they are still in the early stages. They cannot independently complete all tasks and can only excel in certain specific areas, ultimately still requiring human intervention to advance the work to the next stage.
Therefore, we are now in the 'AI-Human Symbiosis Era.' Symbiosis means that AI and humans combine, collaborate, and form a new working model.
The higher your grasp of AI, the greater your value in the workplace.
In the future, a new type of professional will emerge, whose competitiveness comes from a deep understanding of AI and how to utilize AI to accomplish tasks that others cannot.
Another example is the ANA AI developed by Oxford University. This is the world's first AI scientist, which not only can propose new theories and ideas but also can actively test these theories and bring back experimental results.
If it discovers anything new, it can also publish research findings.
The key is that this AI scientist requires data, and it must rely on a large amount of data to function.
In the healthcare industry, there are countless published research papers discussing how various drugs affect the human body, the reactions of patients to these drugs, vitamins, supplements, and so on.
AI can discover hidden correlations from this data. For example, a certain drug was initially developed for a specific disease, but it may actually be beneficial for other diseases as well, and AI can uncover these unexpected effects and publicize them.
This is exactly what is happening right now - AI is helping humans discover new knowledge, new drug uses, and new therapies that we previously did not notice. All of this data has actually existed; it is just that humans have not been able to discover the patterns within.
We are witnessing significant advancements in the field of AI, and these breakthroughs are occurring almost simultaneously.
Jack Dorsey, the person who originally created Twitter, has just launched a new platform called Goose, specifically for building AI agents, allowing any business in the world to easily create its own AI agents.
Moreover, this platform is open source and can be used by anyone. More importantly, it can be used across all major platforms; it does not belong to OpenAI, nor to Google or Microsoft.
This is big news in Silicon Valley.
8. AI's reasoning methods are completely different from humans.
Can AI really reason.
The answer is complex. AI can indeed reason, but its reasoning method is entirely different from ours.
For example. In an experiment, researchers found that when you ask AI a question, simply changing one or two words in the prompt can lead to completely different answers. In fact, just altering a few words can decrease the AI's accuracy by as much as 65%.
What does this mean?
When we reason and draw conclusions, we rely not only on language but also on our experiences in the real world.
Our knowledge is built on real-world experiences. For instance, if someone asks you: "What did the dog bury in the yard?"
Most people would answer: "A bone."
Because we know that dogs like bones, and dogs typically bury bones in the yard. We are able to reach this conclusion because we have a cognition of dogs and understand their habits.
But AI operates quite differently. AI may give the same answer, but its reasoning path is entirely different from ours.
AI does not actually know what "dog" is, nor does it know what "bone" or even what "yard" is. For AI, everything is just mathematics.
AI does not understand the world like humans do through experience; it is purely based on mathematical calculations. It simply analyzes the structure of the sentence "The dog buried ___ in the yard and then fills in the blank based on statistical probability. It is just calculating: in all the text data that has appeared in the past, which words are most likely to appear in this blank?
This is its reasoning method, completely based on probability rather than real-world experience.
So how does AI fill in this blank? It searches through all the data to see which word is statistically most likely to appear, and then fills in the word "bone."
But if you slightly change the way you ask AI, it may give you a completely different answer, because it does not truly understand the question.
This is also the reason for AI's "hallucination"—AI sometimes gives completely absurd answers. Because AI does not understand the real world, it only processes the mathematical relationships between words based on data and statistics.
As humans, we often ponder a question: what is the difference between us and AI? The answer is, we are completely different.
You don't need to worry, AI will never become human because AI cannot experience this world as we do. It knows nothing about the world; it can only process the data we provide.
If AI receives incorrect data, it will output incorrect information, because it fundamentally does not understand the real world; it is just a data processor calculating mathematical equations.
This point is very important.
9. AI deep thinking provides more accurate answers.
The trend we are seeing now is that the longer AI takes to think about a problem, the more accurate the answers it provides.
Therefore, the latest trend in AI processing is to give AI more time to think.
Of course, AI does not "think" like humans; its way involves repeatedly querying different versions of questions on the same topic, generating different answers, and then comparing, filtering, and optimizing to finally provide the best solution.
This is called "deep research." AI will first formulate a plan; after you ask a question, it will develop a research plan, and only after you confirm will it start working.
This may take five minutes to give you an answer, or it may take five hours, or even longer.
Anthropic, invested by Amazon, recently launched a new feature that allows you to specify how long AI should think.
This is the future development direction - the thinking time of AI will be controlled by us, and its length will be determined by us.
There is also Notebook LM, which allows you to summarize various data from the Internet. Once the data is imported into Notebook LM, you can 'converse with the data' and directly ask questions to it. Even more interestingly, you can instruct AI to discuss the data itself, creating two virtual characters to have conversations with each other.
The breakthrough of this technology means that even if you have no understanding of AI and no experience, you can utilize these tools like a data scientist and start using them immediately to optimize your Business.
These products are entering China, and similar technologies are also being developed domestically. So, since you already know these trends, now you can try these tools to see how to use them to enhance your Business capabilities.
Recently, there is a very interesting Venture in Silicon Valley called Gecko, which can automate recruitment interviews.
In the future, many job seekers may face an AI interview before formally communicating with human interviewers. This is the era we are entering.
10. Robotics research is moving towards 'more human-like'.
Faith A. Lee is a researcher at Stanford University, a very smart Chinese scientist who recently founded a new company focusing on spatial intelligence.
What is spatial intelligence? This is a type of intelligence used in the field of Robotics. Making robots smarter is not an easy task, as robots must be able to operate in the real world, rather than just involving language processing like DeepSeek or ChatGPT. Robots need to be able to perceive the real world and interact with it intelligently.
Her research focuses on training AI to 'see' the 3D world from 2D images.
For example, when we see a picture of a cat, we can immediately understand its scene, even if it is a flat 2D image; we can still form a three-dimensional 3D picture in our minds.
For instance, imagine a cat stretching its paw to touch a cup of milk. After seeing this scene, we can quickly predict: 'This cup of milk might be knocked over by the cat, and the milk will spill everywhere.'
But AI doesn't have human experience; how can it infer the development of this scene from a static 2D image?
Faith is researching this problem; she trains AI to predict what will happen next, such as: 'This cup of milk might tip over because it is liquid, so it will flow onto the table and even spill onto the floor.'
This is a very difficult question, but if robots want to operate in the real world, they must possess this ability to adapt to various unpredictable situations.
Apple is also developing robot software, researching how robots, even if they are not human, such as a lamp, a machine, or even a car, can convey emotions through their natural movements.
MIT is developing an AI capable of recognizing human emotions.
Imagine if these technologies are combined, Robotics would be better than humans at instantly recognizing people's emotions and would be able to accurately interpret human feelings.
Moreover, Robotics itself also needs to express emotions. The famous Japanese robot designer Hiroshi Ishiguro is researching how to create more lifelike bionic robots.
In a few years, we will completely be unable to distinguish between a human and a robot.
In the workplace, we will surely have robotic colleagues. At home, we will definitely have robots helping with chores, taking care of children, and even taking on more responsibilities.
Robot spouses - we are moving towards such a future, and this will happen within the lifetime of our generation. We will create robotic companions that humans cannot compete with.
What will happen when that day comes?
This could be a frightening thing. Because when these robots perfectly simulate everything about us, we might forget how to interact with real people.
They will know how to make us happy, know how to make us feel comfortable, and know how to make us happier than any human partner.
But the problem is—they are not real. They will not truly love you, nor will they genuinely care about you.
It's a double-edged sword. We must use this technology with extreme caution, or we might even affect the most basic emotional connections between people.
Especially our children—they will grow up with robotic caregivers. These robots will behave as if they love the children, and the children will believe this. But not just children, adults will also believe that these robots genuinely care for them, even when the reality is different.
Humanity is approaching an unprecedented critical point.
In the course of history, humanity has never faced such a situation, and now we must find a way to utilize this extremely powerful technology in the right way.
11. Robotics + AI, the direction of human society's future.
Now, there are numerous companies in China manufacturing humanoid robots. Youshu may be one of the most well-known companies, but in reality, there are already hundreds of enterprises globally developing these robots.
AI has become strong enough to understand the world, communicate with us, and perform tasks, yet the manufacturing of Robotics remains a more challenging problem.
Manufacturing Robots is much more difficult than creating AI. For machines to operate stably in the real world without frequent breakdowns and at a low cost, this remains a challenge.
But it's just a matter of time—one day, we will completely solve this problem.
So, what happens when the cost of a humanoid Robot is equivalent to the annual salary of a regular worker?
If you buy a Robot, it can: work seven days a week, all year round; operate 24 hours a day without complaining or asking for a raise; won't get sick and doesn’t need leave; no need to pay social insurance or other employee benefits; if dissatisfied, you can simply replace it; and can be upgraded each year to become stronger and smarter.
As a business, would you choose to invest in human employees or invest in Robots?
Almost all businesses would choose Robots.
Everything I just mentioned is to tell you—the era of AI agents has arrived, and the age of Robots is also on the way.
The software we are building is achieving everything that humans can do; almost everything that humans can do, AI can do.
Once hardware is combined with this software, it will be unstoppable, and society will undergo a complete transformation.
In 10 years, maybe 20 years—while we cannot predict the exact time, it is certain that this will happen—there will almost be no traditional jobs left, as Robotics will be able to perform all tasks at a lower cost.
Robotics + AI is the future development direction of human society.
I do not aim to make you feel pessimistic, but I want you to realize that if you are managing your own wealth, you need to understand the future development direction.
All of this will happen in our lifetime, and those who control Robotics and AI will control most of the world's wealth; this is an undisputed fact.
This is why Elon Musk is investing in AI, and companies like Google, Alibaba, Microsoft, and others are investing hundreds of billions into the AI field. They are all smart enough to see the trends that are unfolding and know that it is inevitable.
This is also why every company must consider this trend, as the scale of this transformation will far exceed that of the Industrial Revolution, and its impact will make the Industrial Revolution seem insignificant.
Smart investors do not place all their bets on one technology.
So, what would a world without work look like? This is a difficult question to imagine, as most of our identities are built upon work.
If AI and Robotics are better at our jobs than we are, what do we become?
These are extremely profound questions, and currently, very few people are truly contemplating them. But everyone will eventually be forced to think about it, as we need to understand how to use this technology correctly.
The true purpose of AI and Robotics should be to improve our lives, not to deprive us of our value.
Professor Michael Kozinski from Stanford University is using AI to analyze human facial features and claims that AI can determine your personality, beliefs, and behavior tendencies by observing your face.
He himself is a believer in Feng Shui and even incorporates Feng Shui theories into his system.
This technology is both fascinating and frightening. But regardless, it is being developed, and that is the true power of AI.
Currently, there is a new technology called Teslin Machine, a completely new logical computation technology.
Recently, discussions were held with some top Computer scientists who are researching how to develop a new algorithm that theoretically could increase GPU training speed by 1000 times, much faster than Google’s current use of TensorFlow, while also improving energy efficiency by 10,000 times.
If this breakthrough becomes a reality, the computing efficiency of AI will undergo a disruptive leap.
Of course, this technology has not yet been validated; it is still in the laboratory research stage.
I want to tell everyone that what we consider 'certainty' today may be completely different tomorrow.
The progress of Technology does not develop linearly; it does not grow steadily, but rather explodes.
The development of Technology is like this: steady, steady, then—boom! Suddenly it erupts.
So, when investing in AI, such as buying NVIDIA Stocks, remember that it may be the star of the market today, but it could crash tomorrow.
As I heard the advice before: "Do not put all your investments in one place; diversification is a wise strategy."
Smart investors do not bet everything on one technology; they diversify their investments and place bets on different possibilities.
AGI (Artificial General Intelligence) refers to AI having an intelligence level equivalent to that of humans, capable of thinking and solving problems like humans.
Sam Altman, the CEO of OpenAI, has said, "The age of AGI has arrived, and we are at the doorstep."
My view is slightly different; I believe that when AI has enough data, and that data is correct, it can be smarter than humans and even surpass humans in certain areas.
Currently, we are filling the data gaps. For example, in certain fields, AI is already more powerful than humans, such as: in Medicine, AI is more accurate than doctors in certain diagnoses; in Law, AI is faster and more precise than lawyers in analyzing certain cases.
AI does not think like humans do, but it can provide better answers than we can.
The more data we provide to AI, the broader its range of capabilities becomes, and that is precisely what is happening right now.
13. Remember: AI will never be a true human.
The final question is whether AI can possess consciousness.
My answer is: I don't know.
However, one thing is certain; AI will never have consciousness like humans.
As previously mentioned, AI can never experience the world like we do. It is not biological, it lacks chemicals in its brain, it has no hormones, and it does not perceive emotions like we do. AI will never have true self-awareness like humans.
But AI can act as if it possesses consciousness.
In fact, some advanced AI today are already making it difficult to distinguish them from humans. In ten years, I can guarantee that when you converse with AI, you will not be able to tell if it is human.
The anthropomorphism capabilities of AI will become so powerful that it will even intentionally make mistakes, pretend to forget things, and mimic human thought processes to appear more like a real person.
In the real world, AI will exhibit all human emotions. It may cry, it may say it loves you, and it may do all the things only humans can do.
But it will never be a real human, and we must always remember this.
Editor/jayden