OpenAI has released another heavyweight new product-SORA Wensheng** model! Will AGI implementation be shortened from 10 years to 1 year?
OpenAI has released another exciting news, they have developed a heavyweight product - SORA Wensheng ** model.
According to reports, SORA can quickly generate a minute-long realistic movie texture according to the text instructions entered by the user**. As soon as the news came out, it sparked a wide range of discussions, and many AI experts expressed their opinions.
Tesla's Musk lamented: "Mankind is willing to gamble and lose." Zhou Hongyi, founder, chairman and CEO of 360 Group, predicted: "The realization of AGI will be shortened from 10 years to 1 year. ”
It is undeniable that the emergence of SORA will have a profound change and impact on the profession and society. It presents both challenges and opportunities. In terms of careers, SORA may change the way professions work in production, editing, animation design, etc. While this will reduce the need for traditional production skills, it will also create new career opportunities, such as AI content curation and editing.
It is worth noting that with the continuous progress of the SORA large model, there will also be a greater demand for computing power. Taking OpenAI's official website as an example, since 2012, the amount of computation required to train AI models has doubled every 3-4 months. Taking the GPT-3 model as an example, according to Lambdalabs data, the model has a parameter scale of 175 billion and a complete training operation of 3640 pflop s-days. A single training session of the model takes about 355 CPU-years and costs $4.6 million.
After further data search, it was found that the ** released by OpenAI showed that the SORA large model inherited the technical route of the previous GPT model, that is, "source data - Transformer - Diffusion - emerge". The development of this technical path is mature, and it undoubtedly requires a large amount of data, large-scale parameters, and huge computing power to support it. In particular, since the amount of data required for training itself exceeds the text, it is reasonable to foresee that the SORA model will further exacerbate the demand for AI chip computing power.
The core component of computing power is AI chips, and the increase in computing power demand caused by the development of large models will also stimulate the demand for AI chips, the core component of computing power. According to the data cited in a research report at the end of last year, the global AI chip market size will be about $45 billion in 2023 and is expected to reach $400 billion by 2027, with an average annual compound growth rate of more than 70% in the next five years. At present, the representative manufacturers of domestic AI chips include Cambrian, Haiguang Information, Loongson Zhongke, Suiyuan and so on.
Chip. Summary:
The release of OpenAI's SORA model marks a major breakthrough in AI technology, which accelerates the arrival of the AGI era and allows us to see opportunities for AI chip manufacturers. Domestic AI chip manufacturers should seize this opportunity, increase investment in AI chip research and development, and improve the performance of AI chips to meet the growing market demand. At the same time, domestic AI chip manufacturers also need to pay attention to the international market dynamics and keep pace with the global technology development trend to ensure that their products can meet the needs of the AI chip market at home and abroad.
This article is the original article of Beego Beet Information, and it is not allowed to be ** without permission and authorization.