ChatGPT, stuck in the web, wants to break out in the mobile ecosystem
Since the beginning of this year, major companies have reached new heights on the road to large models. First, domestic companies HE Tuber regardless of size have participated in the development and experimentation of LLM models. Second, even well-known large models such as ChatGPT and newBing have encountered traffic bottlenecks and have stopped growing. In this case, developing mobile terminals and trying new scenarios has become the same choice for everyone.
“I believe that many of us have tried ChatGPT, and I also believe that many people have put it aside after trying it, because it is still basically separated from our work, so we put it down after using it. But I I still hope that everyone will not get up early and rush to the late market, because this is a paradigm revolution that will bring about disruptive changes."
Recently, Wei Qing, chief technology officer of Microsoft China, said the above passage at a new book reading meeting. The second half of the sentence still needs time to be verified, but the "fragmentation" of the first half has become a common reality.
ChatGPT has been on the Apple App Store for one month and has received 5 million downloads. However, its web traffic entrance is gradually slowing down. The previous miracle of rapid growth turned out to be short-lived.
The total number of users is limited, and the faster the growth rate, the shorter the time to reach the ceiling. In less than a year, ChatGPT, which has already encountered a traffic ceiling on the PC side, has targeted the mobile Internet as its next step.
1. The desktop ecosystem has reached its peak
Even if ChatGPT soars like a rocket, it will eventually face the law of gravity.
According to data from the analysis agency SimilarWeb, ChatGPT's web page visits began to experience a slowdown in growth. Although global visits still reached 1.8 billion in May, the month-on-month growth rate was only 2.8%.
Compared with the three-digit month-on-month surge that started in January, the concept of large models with reality-distorting force fields in the past is returning to the physical world.
This is not only a problem ChatGPT itself is facing, but Bing, which has the same large model core, is also returning to the old track.
Data insights from SimilarWeb show that Bing, which introduced the GPT-3.5/4 model in February, among the natural search keywords from March to May, the natural traffic entered by users using keyword searches such as New bing, which represents the large model version, totaled A sharp decline of 56.84%
On the contrary, the natural traffic keywords related to traditional search engines are returning, which to a certain extent means that search engines empowered by large model capabilities are not that easy to use.
Taking market share as an example,
The growth expectations of the large model represented by New Bing have continued to fail. The market share in March this year was 2.86%, and fell to 2.77% in May, returning to the normal market fluctuation range.
At the same time, competitors of the same type began to copy ChatGPT’s initial growth rate and catch up.
For example, Character.Ai, the second largest artificial intelligence chat website, had 281.4 million visits in May, a month-on-month increase of 62.5%; Google Bard, which was launched in February this year and fully opened in early May, had 142.6 million visits in the same month, a month-on-month increase of 187.2% %.
Although the two are still far behind the 1.8 billion visits of ChatGPT, in the competitive ecosystem of PC desktops, ChatGPT is obviously facing the hunting of more latecomers.
New opponents emerge one after another, and old partners are also causing trouble. ChatGPT, which "grabs" food with its subversive user experience, is an out-and-out barbarian in the eyes of traditional websites.
In May alone, according to incomplete statistics reported by
The media, three companies were the first to come under pressure due to the traffic growth of ChatGPT.
First, Reddit, the largest social news forum in the United States, experienced a decline in traffic, and then announced plans to charge for APIs. As OpenAI’s earliest source of large model training data, Reddit is trying to make up for lost business interests in this way.
Coincidentally, Stack Overflow, one of the most commonly used websites by programmers, also encountered the same dilemma.
Network analysis company SimilarWeb reported in April that since March this year, Stack Overflow's traffic has dropped by 13.9% compared with February, and the website's traffic in April has dropped by 17.7% compared with March.
From long-term single-digit fluctuations in 2022 to double-digit declines for many consecutive months this year, this is mainly because more users use ChatGPT to help debug code. Programmers are currently the segment with the highest usage rate and frequency of ChatGPT.
Chegg, a vertical academic forum website, was previously very popular among college students. Because it can help students complete complex homework assignments, such as those encountered in STEM courses. Like Stack Overflow, users post their homework questions and then wait for a human response.
However, this business logic is facing pressure from ChatGPT, which can provide students with instant answers. Chegg recently reported on its earnings call that the company's stock price and revenue were declining, and that ChatGPT was the primary reason.
Dan Rosensweig, CEO of Chegg, said on a conference call earlier this month: "Since March, we have seen a significant increase in student interest in ChatGPT. We now believe it is having an impact on our new customer growth rate.
"
In the PC desktop ecosystem, facing the pressure from ChatGPT, Stack Overflow and Chegg have successively announced measures such as API payment and the development of large models in vertical fields to raise the threshold for desktop data circulation. This is also the current situation of the entire industry.
Take the domestic market, which is the most active for large models, as an example. According to the "China Artificial Intelligence Large Model Map Research Report" released in May by the China Institute of Scientific and Technological Information, a subsidiary of the Ministry of Science and Technology, as of May 28, at least 79 domestic basic large-scale models with a scale of more than 1 billion parameters have been released.
However, according to Caijing, many industry insiders believe that there are very few companies that have the actual ability to commercialize large-scale models. When the growth rate of ChatGPT, the industry leader, slows down, large models are also coming to the end of PC desktop product release: breaking away from the early market novelty and entering a new node of implementation in a more competitive environment.
An American analyst commented: "In the history of technology products, there has never been a lack of epoch-making technological novelties, such as the first year of VR, the first year of autonomous driving, and the first year of artificial intelligence. But most of the time, these The so-called first year is ultimately the end of the early adopter mentality that has swept users around the world, and they are waiting for the next restart.”
For ChatGPT, after sweeping users around the world with its large-scale technology experience, although it is not so pessimistic to wait for a restart, it is urgent to find more scenarios to retain users.
