Generative AI is in vogue. But this technology has great obstacles. Due to the fact that they classify a large volume of existing content to create something new, they require enormous computing power, delivered through the cloud. Y encontrar a alguien con un servicio a la altura que permite esto es muy complicado. But for companies, this need is, without a doubt, a golden thread.
Although Microsoft, Google, OpenAI and others have their own automatic learning models to process and generate text, audio and images, Amazon aún carece de ese frente, but what it does have is one of the largest cloud computing networks in the world. So, to overcome this problem, Amazon Web Services (AWS) has announced a long-term strategic partnership similar to Hugging Face.
For those who don’t know, Hugging Face is one of the central points of automatic learning, with more than 100,000 accessible and free models downloaded more than 1 million times a day by researchers, data scientists and engineers. AWS, for su lado, es one of the major cloud computing providerswhich offers tools to help developers create software based on IA, including patented computer chips that transmit large amounts of data and drastically reduce the time it takes to create a chatbot or other IA products.
Through this partnership, this AI service will be able to access AWS tools (Amazon SageMaker, AWS Trainium, AWS Inferentia, etc) to train, adjust and implement next-generation artificial intelligence models in AWS. The objective, they say, is to make automatic learning models more accessible to the community with greater performance at lower cost.
The war to lead from the market
Amazon also wants, of course, that this alliance serves to plant the face of ChatGPT, which has assumed an antes and a después in the vision that the world of generative AI has. All this while more and more technology companies are blinded by allies in a booming market.
And, as this sector grows, it is more important than ever to guarantee that all developers can access and evaluate the latest popular models of generative AI. Sin embargo, most of these are not publicly available, which widens the gap in automatic learning capabilities between the largest technological companies and all others. AWS and Hugging Face want to counteract this tendency and, according to dicen, democratize automatic learning.
In addition, Hugging Face has announced that it will build the next version of the BLOOM language model in AWS, according to Swami Sivasubramanian, vice president of database, analysis and automatic learning in Amazon’s cloud unit. Specifically, the next generation of Bloom, an open-source AI model that competes in size and scope with the model that Microsoft-backed OpenAI used to create ChatGPT, will run on Trainium, a patented artificial intelligence chip created by AWS .
Hay que commentar que esta association es parte de a growing trend of deals and investments which binds the largest cloud providers with companies that work in generative AI. El mes pasado, Microsoft Corp. reached an agreement to invest in the maker of ChatGPT, OpenAI, which is said to be worth 10,000 million, and is using the technology of the startup for Bing search.
At the beginning of this month, Google also invested almost 400 million in OpenAI’s rival, Anthropic. As we can see, the ability to sell computing services in the cloud is at stake to take advantage of the boom in interest in AI generative programs. Y nadie quiere perderse este for a moment.