Recently, nearly a year after the launch of ChatGPT, the first OpenAI Developer Conference (OpenAI DevDay) was held in San Francisco, and in just one hour released three major blockbuster news, including the GPT model iteration, the introduction of AI personal assistant, and the upcoming launch of the GPT store.

According to OpenAI CEO Sam Altman, OpenAI will launch the GPT-4 Turbo, which is faster, stronger, and cheaper. Specifically, OpenAI will charge based on the total number of token inputs and outputs called by the API, allowing developers to call more tokens and correspondingly reducing costs-the input cost of tokens is 3 times cheaper than GPT-4, while the output cost is 2 times lower.

in addition, OpenAI also announced the introduction of a "copyright shield" mechanism at this conference. when ChatGPT enterprise users and API users are involved in copyright lawsuits, OpenAI will come forward to defend and bear the resulting liability for compensation. This is in line with Microsoft, Adobe, etc.

It is worth noting that OpenAI has also released a special AI API assistant (Assistants API), which provides functions such as code interpreter, retrieval and function call. A lot of the work that developers had to do on their own can now be done by Assistants APIs, making it easier to build assistive AI applications.

GPT-4 Turbo sale

The relevant analysis points out that the more cost-effective large model radiates the entire AIGC application-side industry chain, or benefits related enterprises.

Large model performance optimization is the core of AIGC application experience optimization, and large model call cost reduction is the premise of AIGC business model maturity. With the sharp price reduction of GPT-4 Turbo and the liberalization of Assistants APIs, AIGC product prices are expected to fall and user payment penetration is expected to increase, thus helping the rapid promotion of AIGC applications, accelerating the commercialization of AIGC applications and user consumption mental development.

At present, domestic AI industry leaders are accelerating the expansion of the scope of application landing. Take AIGC Software A- share listed company Wanxing Technology (300624.SZ) as an example. Since the ChatGPT detonated the AIGC singularity last year, Wanxing Technology has launched a number of new AIGC applications including Wanxing Love Painting, Wanxing Broadcast Explosion, Wanxing Smart Performance, Wondershare VirtuLook, Wondershare Kwicut, etc., covering multi-modal AI generation capabilities such as AI picture generation and AI video generation. In addition, Wanxing Technology's original creative software products Wanxing Meow Shadow, Wondershare Filmora, Yitu Diagram, Yitu Brain Diagram, Wanxing PDF, Modao Overseas Edition Mockitt, etc. have all integrated AI functions into the products.

it is worth noting that not long ago, Wondershare Filmora and wanxing meow yinghuan released a new version, which integrates generative AI and interactive AI. it launched AI copywriting, AI music generator and AI text forming three multimedia content AI intelligent generation functions at one time, as well as AI text quick cutting, intelligent voice separation, AI intelligent mask and AI intelligent frame filling four AI intelligent editing functions, and launched the world's first video creation software AI assistant Copilot, to promote video creation into a new era of AI empowerment.

Wondershare Filmora and Wanxing Meow Shadow Global Rejuvenation

in addition, Jinshan office (688111.SH) has also become a member of domestic AI application enterprises worthy of attention with its WPS AI. Not long ago, Jinshan Office and iFLYTEK announced a strategic cooperation. On the basis of existing cooperation, the two sides will continue to give full play to their respective technical and product advantages, and carry out in-depth cooperation in digital office products, industry solutions, AI large models, open capability integration and other dimensions, so as to promote the integration and innovation of technology, products and solutions, and jointly build a digital and intelligent service ecology.

previously, jinshan office announced in September that WPS AI, an intelligent office assistant based on a large language model, had been connected to all products of jinshan office. at the same time, it showed the latest progress of jinshan office self-research model for the first time. The model is based on an open source base, and through training tuning, it helps WPS AI to become the first AI office application in China.

disclaimer: market risk, choose carefully! This article is for reference only and is not a basis for sale.