Produced by|Sanyan Pro Author|DorAemon
In the early hours of this morning, OpenAI announced the launch of GPT-4o. The model is an updated version of the GPT-4 model and will be available to free customers.
According to reports, GPT-4o ("o" stands for "omni"). It can implement any combination of text, audio, and image as input, and generate any combination of text, audio, and image output. And there are many amazing applications:
This is indeed a great update and upgrade that pushes AI technology further.
However, what I want to talk about today is not how powerful ChatGPT-4o is, but the difference in the attitude of some domestic self-media/media and foreign media on this matter.
Netizen puzzles
Yes, you guessed it right, as soon as ChatGPT-4o was released, various self-media and circle of friends "exploded" again! Shock! Turn the world upside down! "That's it.
Every time a new technology comes out, the circle of friends and the platform are filled with all kinds of terrifying titles of "xxx subverting the world, exploding, and blockbuster", but it doesn't take long for everything to calm down again, as if nothing has happened.
So, is ChatGPT-4o so "explosive" this time, and what is the attitude of foreign media to report on this matter?
Foreign media: Neutral introductions are the mainstay, and the title is not "explosive"
Since ChatGPT-4o is a product launched by an American company, let's take a look at the attitude of foreign media first.
To be honest, the author turned many pages of foreign media news, and only felt two words: boring.
Because the vast majority of foreign media are very neutral to the title of ChatGPT-4o, there are no subjective words such as "amazing", just a normal expression OpenAI launched a new version today.
Looking at the content of foreign media reports, the author checked dozens of articles and found that all foreign media reports on ChatGPT-4o objectively describe the performance and role of this large model, and do not mention the argument of "how to subvert the world".
Domestic Internet: Exploding, Subverting the World, and Bombing the ......
Looking at the headlines of domestic media reports on GPT-4o, it is "rewriting history, the most powerful in the world, blockbuster release, a small step for OpenAI, a big step for human AI assistants, all-rounder, subverting the world, a late-night bomb, king bombing, completely changing humanity".
In terms of content, these articles with sensational titles are actually simply about the capabilities of ChatGPT-4o and what kind of functions it can achieve.
But in the end, there is no relevant description of how it "exploded and subverted the ...... world".
But in the same media, in the articles that introduce domestic large models, it is rare to see the title "shocking body", and the content is much more serious, which is obviously different from the attitude of introducing OpenAI products.
SORA also "exploded" after its release
The same is true of the Wensheng video model SORA launched by OpenAI before, which is "explosive, explosive, and dry Douyin" on the domestic Internet; However, in the headlines of foreign media, there are few emotional expressions in terms of various "mediocre".
If you only look at domestic Internet content, whether it is SORA or ChatGPT-4o, it seems that our lives have been completely changed in an instant, however, the fact is that before SORA did Douyin, it was revealed that there were artificial special effects.
So, is ChatGPT-4o, which has been blown into the sky by domestic self-media, really "exploded"?
The attention is not very high
Today, Zhu Xiaohu, managing partner of GSR Venture Capital Fund, commented that the iteration curve of large model technology has slowed down significantly; The start of the volume is free, indicating that GPT user and revenue growth has reached a bottleneck, and model companies that are not deeply bound to large manufacturers are basically out. Third, applications will explode rapidly, and if the cost of inference drops by an order of magnitude, national applications in the AI era will be popularized.
Li Zhifei, founder and CEO of Mobvoi, said in an article today: "The scumbag of human-computer interaction, because of the emergence of GPT-4o, is expected to really be a new person, start the next wave of technology, applications, and business model revolution, and look forward to the next decade." ”
The evaluation of these two, if you taste it carefully, there is no "sense of explosion".
In addition, what makes the author find interesting is that when I read the opinions of netizens on ChatGPT-4o today, I found that the attention was not as high as I imagined, at least not to the extent of "explosion".
Many bloggers posted about the ChatGPT-4o experience being very good and powerful, but the number of comments was very few; Some netizens even ridiculed "changed the world several times", and it seems that more than one person feels the same way as the author.
Not long ago, when SORA came out, there was a lot of discussion on topics and blog posts related to AI, and the popularity greatly exceeded that of ChatGPT-4o.
GPT-4o really subverts the world, is it Wang Bang?
Objectively speaking, GPT-4o is richer in features and more powerful.
For example, it can help children with homework;
helping the blind to take a taxi, etc.; But it is too early to say that the so-called "subverting the world, exploding, and changing mankind" is a bit premature.
AI technology is still developing, and even though it has made impressive achievements and can improve the productivity of some people, it will take time to completely change the world and change everything.
Just as when the electric light was invented, it did not immediately push mankind into the era of the electric power industry, but also needed continuous improvement and popularization before it could be applied.
The headlines of the domestic media that "explode and subvert the world" at every turn have the element of "headline party", which is the result of the pursuit of clicks and readings.
Looking back on similar events in the Internet technology circle in the past few years, from blockchain, to the metaverse, to large models, every time without exception, the "big change" is coming, but in fact, our lives have not changed so quickly, right?
Technically speaking, a deep AI user told me that GPT-4o is not explosive or subversive, because Google's Gemini also has related functions, and document, audio, picture, and video analysis are also very powerful, of course, 4o's reasoning ability this time is much stronger than before.
Another friend talked about the criteria for exploding, one is to arouse attention and discussion in different fields, and the other is to focus on unconventional impressions or original boundaries. The release of 4O, in fact, is still the ability to analyze multimodality, which is manifested in the faster response speed of input and output, close to human, and it is difficult to say that it has broken through the original boundary.
The main reason is that the frequency of explosions is too high, and they don't mean to belittle the relevant media, they are still very professional. is also very dedicated, starting at 1 o'clock in the morning to guard the press conference, writing scripts all night, very hard. When you wake up in the morning, their pictures, texts, and videos have already been rushed out.
Of course, everyone has a different understanding of the explosion, and there is no right or wrong. If there was a level, would the order of the three words "amazing->explosive->subversive" be more reasonable?
In the early hours of the morning, Google will also hold a developer conference, and AI is also the focus, do you think it will explode?