chadbrownpokerplayer| Google and OpenAI suddenly turn focus to AI smart assistants

Every reporter, Yang, every editor, Zhang Haini

It's not what people expect.ChadbrownpokerplayerAI search, the focus of the competition suddenly turned to the AI intelligent assistant.

Recently, OpenAI launched the omnipotent model GPT-4o, which can accept multiple inputs and generate corresponding outputs, demonstrating the new capabilities of millisecond response and multimodal interaction. Meanwhile, Google unveiled its AI assistant Astra and its flagship big model Gemini at its Iamp O developer conference.

Some industry insiders believe that OpenAI did not achieve the breakthrough expected by the market this time, but integrated the existing technology. In addition, Google's layout and innovation in the field of AI search, as well as its efforts in smartphone operating system optimization, show its deep accumulation and strategic layout in the field of AI.

Competition to the white-hot stage, it seems to have bid farewell to the simple technology competition, but also the application and user experience competition. When the influencing factors become complex, what are the chances that OpenAI, which is focused on the cutting edge of the big model, will become a winner?

Raiding Google, OpenAI announced AI personal assistant first.

The expected "war" around AI search did not begin, and the focus shifted to the AI intelligent assistant.

OpenAI unveiled its latest product GPT-4o, "o" or omni, meaning "omnipotent", at a press conference the day before Google's IZP O developer conference on May 13. According to the OpenAI website, GPT-4o is a step towards a more natural human-computer interaction because it accepts any combination of text, audio and image as input and generates any combination of text, audio and image output.

Mira Murati, OpenAI's chief technology officer, said at the launch that GPT-4o is twice as fast as the existing GPT-4 Turbo, but at half the cost. GPT-4o can reason text, audio and images in real time, and the response time almost reaches the human level.

During the 26-minute live broadcast, GPT-4o demonstrated a series of new capabilities, such as millisecond response, recognition of human emotions for audio and video interaction, and multimodal input / output. At the same time, GPT-4o covers desktops and App and is completely free to users.

Google, on the other hand, demonstrated its omnipotent AI capabilities at its Iamp O developer conference, where it released and updated more than a dozen products, including AI assistant Astra, Vincent model Imagen3, Vincent video model Veo for standard Sora, and flagship Gemini.

In Google's demo video, when using AI Assistant Astra, as long as you turn on the phone's camera and aim at any object, AI can accurately name the item. By pointing the phone's camera at an object, Gemini can recognize it, such as a red apple, and answer questions such as "what can be heard in the lens?"

In addition, Google said it will expand Gemini's multimodal features in the summer, including the ability to conduct in-depth two-way conversations with voice, a feature known as Live. With GeminiLive, users can talk to Gemini and choose which sound it responds to from a variety of natural sounds. Users can even speak at their own pace, or interrupt and clarify questions during answers, just as in any human conversation.

Recently, it has been revealed that Apple is finalizing an agreement with OpenAI to introduce some of the latter's technology to iPhone this year, Bloomberg reported. At this press conference, Google Vice President of Product Management Sameer Samat made it clear that Google will further optimize its Android operating system through Gemini. This optimization will first be reflected in Google's own mobile phone Pixel.

GPT-5 is absent, OpenAI slows down?

For the surprise update of OpenAI, the industry is no longer a unified admiration. "although the press conference is amazing, Google should not panic after reading it." Fu Sheng, chairman of Cheetah Mobile and CEO and Orion Star, said on his personal Weibo.

In a short video released on May 14, Fu Sheng said that "all domestic artificial intelligence practitioners stayed up late waiting for the release of the 'nuclear bomb' on the other side of the ocean, but did not expect that the 'nuclear bomb' was not released, but pulled out a bunch of 'drop guns'." He said it was disappointing that OpenAI did not release GPT-5 this time.Chadbrownpokerplayer.0, even GPT-4Chadbrownpokerplayer.5 did not see, but instead released GPT-4o, "is to combine a series of engines, such as pictures, text, sound, so on."ChadbrownpokerplayerYou don't have to switch back and forth.

However, Fu Sheng also said: "OpenAI this time in order to allow more users to use it, it can be said that riveting is full of strength, a series of applications, API price reduction, GPT free." We certainly hope that OpenAI can make the industry develop better, and we can also learn seriously. This conference really tells us that applications are promising and that everyone should work hard. "

"when GPT-4o comes out, it is much more advanced than before. Every time it (OpenAI) upgrades, it will 'die' some companies. This time, some teams that do GPT real-time voice interaction can directly announce that they have been disbanded." The day after GPT-4o was released, a big model industry entrepreneur lamented to the Daily Economic News.

Shenyang, director of the Yuan Universe Culture Laboratory of the School of Journalism at Tsinghua University, also tried GPT-4o the next day, saying in its video number that at the textual level, GPT-4o is more excellent in hair fineness, light and shadow effects and other details.

With the encounter between the two sides on smart assistants, Shenyang believes that the current competition situation has become clearer, Google is further promoting its Gemini-based AI assistant, and Apple and OpenAI have also reached a preliminary partnership to install ChatGPT on Apple phones.

Shenyang said that with the launch as a turning point, ChatGPT has been transformed into a soul mate, so the industry is clear that Apple uses built-in ChatGPT against Google Gemini's mobile assistant. Meta will also launch a Llama-based mobile assistant. For the industry, AI assistant is expected to move from 100 million users to 1 billion users.

"GPT-4o, when it was released, I thought it was very powerful, but now I think Google's latest product has completely caught up with its achievements. I think OpenAI should be more nervous later, because application companies and super platform companies are all catching up, and its advantages are getting less and less." Li Mingshun, founder of Shun Fu Capital and chairman of Bank AI, told the Daily Business News that at present, the user growth of OpenAI is not obvious, technology leadership and cost advantages are not necessarily the best, at the same time, the era of strong applications is coming faster and faster, under this background, Google has put all its applications together with large models to form a stronger user dependence, or better.

In Li Mingshun's view, in the next stage, platform application companies in the United States, including Microsoft, Apple, Dell, as well as China's Tencent, Byte, and Ali, will successively combine their own application and large model capabilities to launch super applications, gradually moving into an era of comprehensive competition, and it will be more difficult to rely on a big model.

The search battle is not over, Google counteracts "encirclement and suppression" OpenAI

Eating the giant's "cake" is not that easy. Before the launch, the market was filled with OpenAI's smoke bomb around the search layout, and it was reported that OpenAI was likely to launch a new search engine based on ChatGPT technology. At the same time, a web page called "GPTSearch" has been launched, but currently only members can access it. PeteHuang, a well-known journalist, had previously tweeted that GPTSearch would be officially launched on May 9.

In the end, Google held its ground in this round. Google CEO Sandal Pichai (Sundar Pichai) mentioned in his speech that one of the most exciting changes brought about by Gemini is that in Google search, "one of our biggest investment and innovation areas is our founding product-search."

From the launch, Google has taken the lead in combining AI capabilities with its search engine. Google announced that the "AI Overview" (AI Over-views) function, which can summarize Google's search engine results, will be launched in the United States this week. In this feature, Google will show users the answers generated by AI.

According to Google, AI overview is designed to respond to more complex searches and help users find solutions. For example, when people search for vegetarian preparation or travel plans, the answer provided by AI appears at the top of the search page.

Google has also improved the visual function of search to support asking questions through video. At the AI O developer conference, Google demonstrated that when faced with a record player failure, users can shoot a video and ask questions, and get an overview of Google through a new search, including repair steps and resources.

Despite sniping at Google's new products at the smart assistant level, there has been no further action on the much-anticipated search product, a revolutionary feature that has been highly anticipated since the launch of GPT-3.5, after a burst of smoke bombs. In the search market, Google's basic disk is still solid, while its full AI capabilities are vaguely encircling competitors.

On the other side of competing for territory with the giants, the hidden troubles remain for OpenAI.

Just a day after the release of GPT-4o, Ilya Sutskever, co-founder and chief scientist of OpenAI, who had long disappeared from the public eye, announced that he had left OpenAI. Last November, OpenAI management turmoil, Sutskever is seen as the driving force behind the storm. Not long ago, Andrej Karpathy, one of the founding members of OpenAI, also left on the eve of the release of Sora.

In this brand-new technological revolution, as the focus of the competition shifts from the large model technology to the application side, and the OpenAI, which once led the way with a "dark horse" posture, begins to slow down, a new turning point may have emerged.

Tags：

Article Copyright Statement: Unless otherwise specified, all articles are jiliasia16 com original. Reproduction or reproduction should be in the form of hyperlinks and indicate the source.