The enthusiasm for large-scale model entrepreneurship is still increasing, and the opportunities for AI entrepreneurs in the "Hundred Models War" are more at the application layer. In order to lower the start-up threshold for AI entrepreneurs as much as possible, Amazon Cloud Technology and Zhen Fund, a leading early-stage investment institution, jointly launched the "AI Supernova Project", providing AI application entrepreneurs with dreams from cloud resources, model selection, Technical support such as fine-tuning and application landing, all-round support from 0 to 1 product prototype construction, financing acceleration, etc., help early entrepreneurs to start quickly and discover [AI supernova]!
Amazon Cloud Technology Helps AI Entrepreneurs Start Fast
For more than 20 years, artificial intelligence and machine learning have been the focus of Amazon's attention, and more than 100,000 customers have realized innovation with the help of Amazon cloud technology AI and machine learning services. In order to better help enterprises and developers innovate rapidly in the generative AI boom, Amazon Cloud Technology has also launched a series of products and services to lower the threshold for them to build applications based on large models.
Now generative AI has become a hot word among the people, and the explosion of AI technology directly promotes the growth of demand for computing power resources. In response to the blowout demand for computing power, in addition to NVIDIA, Intel, and AMD chip products, Amazon Cloud Technology also provides self-developed general-purpose chips Graviton, dedicated AI chips Trainium and Inferentia, and Nitro System, providing users with a variety of cost-effective choices.
Through the fully managed basic model service Amazon Bedrock, entrepreneurs can more easily build and expand generative AI-based applications, and can access AI21 Labs, Anthropic, Cohere, Stability AI, and Amazon Cloud Technology’s own AI through a simple API interface. Various advanced base models. The new Amazon Bedrock Agents feature extends the underlying model to understand user requests, break down complex tasks into steps, conduct conversations to gather additional information, and take actions to satisfy user needs.
In Amazon SageMaker JumpStart, the machine learning center of Amazon Cloud Technology, you can fine-tune and deploy Meta AI's Llama2, Stability AI's Stable Diffusion, AI21 Labs' Jurassic-2 and other popular large models with one click. You can also seamlessly integrate the open source platform HuggingFace through SageMaker The models in are used for training and inference.
In addition, Amazon CodeWhisperer allows developers to simply provide natural language and let the system generate the code they need.
At the Amazon Cloud Technology New York Summit, Amazon Cloud Technology released seven new functions for generative AI, allowing entrepreneurs to focus on the business side and how to meet the needs of target users, and reduce the cost of generative AI applications as much as possible. threshold.
Entrepreneurial opportunities for generative AI lie at the application layer
Since its launch in November 2022, the number of users of ChatGPT has rapidly exceeded 100 million. The total number of users of Midjourney and Stable Diffusion has exceeded 15 million and 10 million respectively. Various other generative AI applications have also sprung up like mushrooms. This means that the successful "breaking circle" of generative AI has entered the lives of the mainstream public.
According to a Goldman Sachs report, generative AI can boost global GDP growth by 7%, or nearly $7 trillion in increments. Many technology leaders have also asserted that generative AI has the potential to become a bigger opportunity than the Internet. Just like the previous "Internet+", "AI+" is expected to be deeply integrated with all walks of life and continue to bring new application forms. So in the highly competitive ecology of generative AI, where are the opportunities for entrepreneurs?
In the generative AI ecosystem, the infrastructure layer includes chip manufacturers such as NVIDIA, data centers, and power sources, etc., and the software layer has cloud computing companies like Amazon Cloud Technology to provide support for generative AI applications. After all, the large model layer is a giant company or a pioneer game like OpenAI with the full support of a big factory behind it. For early entrepreneurs who lack resources, infrastructure and large models do not seem to be good entrepreneurial directions.
Dai Yusen, managing partner of ZhenFund, believes: "The model itself is an infrastructure that needs to establish a relationship with end users through applications. Developers still need to use various tools to develop applications. Therefore, entrepreneurs with limited resources For example, the application layer (including the developer tool layer) has a lot of entrepreneurial opportunities."
"AI Supernova Project" Recruitment Open
Changes in the field of generative AI are almost daily, requiring developers to build and innovate fast enough. At the beginning of the "AI+" era, Amazon Cloud Technology believes that the most exciting and innovative large-scale model-based applications are still on the way, and is willing to work with its partner ZhenFund to find and support these developers with ideas.