Artificial intelligence and intelligence are not the same thing. Although there is some overlap between the two, they are fundamentally different. Artificial intelligence primarily serves as a tool to help humans solve specific problems (such as playing chess, writing, assisting in analysis, etc.), while intelligence encompasses not only the ability to solve problems but also the ability to discover them (of course, in many specific fields, many people lack both the ability to discover and solve problems, yet they still possess intelligence). However, many people often intentionally or unintentionally confuse artificial intelligence with intelligence.
Artificial intelligence refers to a technology that simulates human intelligence through computer programs. It involves using algorithms and data to enable computer systems to automatically perform complex tasks, possessing a certain degree of perception, understanding, learning, and decision-making capabilities. Intelligence refers to the cognitive abilities possessed by biological entities (such as humans), including perception, understanding, learning, reasoning, and problem-solving.
Artificial intelligence is a technology designed and developed by humans. It is a branch of computer science that achieves intelligent functions by writing code and algorithms, and its capabilities can be customized based on tasks and training. It can excel in specific areas but may be relatively weak in others. Artificial intelligence can extract patterns and rules from large amounts of data through technologies such as machine learning and deep learning, and can self-optimize and improve. However, it has not yet reached the level of consciousness and subjectivity; it operates based on algorithms and program execution.
Intelligence, on the other hand, is a product of the evolution and development of biological entities in nature and is a capability of the human brain. Intelligence is a broad cognitive ability that includes various aspects such as knowledge acquisition, processing, and application, as well as adaptability and learning capabilities. Intelligent agents learn through perception and accumulated experience, and can expand and apply knowledge through reasoning and induction, possessing subjective experiences and states of consciousness.
In summary, artificial intelligence is a technology developed by humans to achieve specific tasks by simulating certain aspects of human intelligence. Intelligence, however, is a broader cognitive ability inherent in biological entities, encompassing various aspects including perception, understanding, learning, reasoning, and problem-solving. Although artificial intelligence has achieved significant success in certain tasks, there remains a noticeable gap compared to human intelligence. Viewing artificial intelligence as intelligence can generate high social attention and excitement, accelerating progress in some areas, but it can also lead to phenomena that recur periodically.
ChatGPT Visits May Experience Negative Growth Amid Concerns of “AI Bubble”
China Central Television Finance China Youth Network 2023-07-03 14:05:02
According to website traffic monitoring service Similarweb, the month-on-month growth rate of ChatGPT visits in January this year was 131.6%, but it has decreased month by month since then. By May, this figure had dropped to 2.8%, and negative growth is expected in June.
Not only ChatGPT, but mainstream AI chatbot websites are also experiencing stagnation or even a decline in traffic and subscription numbers. Some media have pointed out that the initial development logic of AI technologies like ChatGPT, based on large language models, was “the bigger, the better”—larger model parameters and more training data. However, this has now become the biggest bottleneck for upgrading this technology. The parameter count for ChatGPT-3 was 175 billion, and it rose to about 1 trillion for GPT-4, causing development costs to skyrocket from approximately $4.6 million for ChatGPT-3 to $100 million for GPT-4. AI research institutions predict that every 6 to 10 months, the computing power required for an advanced large language model will double, leading to significant increases in hardware investments. Additionally, experts have pointed out that by 2026 at the earliest, the quality data and content needed to train large models may be depleted.
Moreover, the safety applications and regulatory issues of AI technology also urgently need to be addressed. The uncertainties in these areas have given investors more reasons to remain cautious about AI investments. Concerns about the transformation of this year’s “AI boom” into an “AI bubble” have also been growing. Since the beginning of this year, leading tech stocks in the U.S. stock market have seen an average increase of over 50%, far exceeding the gains of other S&P 500 index components. American economist David Rosenberg recently stated that the current AI investment frenzy reminds him of the internet bubble of the early 21st century, as the technology stocks’ weight in the S&P 500 index has reached 27%, consistent with the level before the internet bubble crisis. Wall Street insiders have warned that many AI concept stocks, including those of major tech companies, are already overvalued, and their potential downside risks are worth monitoring.
Chief Investment Strategist at a U.S. investment consulting firm, Lance Roberts: The valuations of major tech companies in the U.S. stock market are indeed very high. Valuations can tell you a lot about future returns over the next decade. However, no one is concerned about valuations right now; the only concern is whether one can make money now, as we have turned the market into a “casino.”