Generative AI is accelerating its implementation: Industry application innovation is ushering in the 'cloud moment'

From the emergence of large models, to the rapid development of computing power and storage infrastructure, to the commercial innovation and application of generative AI, the "trilogy" of the generalization process of artificial intelligence has become a common thread. The main theme of 2023.

According to the latest report released by IDC Consulting, more than 87% of industry users around the world have begun to apply and deploy generative artificial intelligence, and the proportion in the Chinese market is as high as 93%. This shows that generative artificial intelligence is accelerating from the strategic planning stage to the implementation stage, and application innovation in all walks of life has ushered in an opportunity to explode

From an eye-catching perspective, "killer applications" for individual users have high hopes, but the actual results are always difficult to meet expectations; if you return to normalcy, you will find that many industries in the ToB field are generative AI On the main battlefield, some heavyweight innovative applications have quietly sprouted.

If you expand your vision from applications to the entire industrial environment, it is not difficult to find the underlying logic for the accelerated penetration of generative AI in various industries. In the period of transition from informatization construction to digital transformation, the emergence of cloud computing has provided a new evolutionary path for many enterprises that are restricted by IT investment capabilities and have difficulty in promoting business upgrades. The platform features of large-scale operation and elastic scalability provide a new path for complex applications. Innovation escorts; when digital transformation moves into deep waters, a new wave of intelligence is coming, and mainstream cloud vendors also play a pivotal role. It is time to comprehensively reconstruct the base of generative AI.

As the pioneer and leader of global cloud computing, Amazon Cloud Technology has been very good at the popularization stage of cloud computing, allowing almost all industries to "do it again." In the era of generative AI, Amazon Cloud Technology is still the pioneer, and the "cloud moment" for various industries to be "do it again" by artificial intelligence has arrived.

Chen Xiaojian, General Manager of Amazon Cloud Technology Greater China Product Department, explains the big announcement

Recently, 2023 Amazon Cloud Technology re:Invent made the strongest noise - launching a number of major releases around the themes of reconstructing cloud infrastructure, reconstructing computing, reconstructing storage, and reconstructing enterprise-level generative AI. Help cloud customers quickly achieve digital transformation and increase the speed of enterprise generative AI innovation. It is worth mentioning that the 2023 Amazon Cloud Technology re:Invent China City Tour - Beijing Station event will also be held recently, and the "seeds" of commercial application innovation are expected to take root.

By systematically sorting out the latest strategies, products, solutions and application cases released by Amazon Cloud Technology, we can paint an industry picture of the accelerated penetration of generative artificial intelligence. We are full of expectations for the innovative energy released by "Cloud Moment"

Reshape the generative AI application innovation base based on three-layer architecture

can be rewritten like this: To a certain extent, cloud computing and generative artificial intelligence are a mutually reinforcing and interdependent relationship: on the one hand, the cloud platform provides the best platform for the application innovation of generative artificial intelligence. ;On the other hand, generative artificial intelligence also provides a rare opportunity for the continuous upgrading of cloud computing

Although it is too early to say that cloud service providers will fully invest in generative artificial intelligence, judging from the latest strategies released by several mainstream cloud service providers this year, most of them will focus on generative artificial intelligence. , the infrastructure, product solutions and cooperation models have all changed as a result

In the field of generative AI, the overall layout of Amazon Cloud Technology can be divided into three levels. The first is the application layer built using the base model, followed by the tools layer built using the base model, and finally the infrastructure layer for base model training and inference. At the re:Invent conference in 2023, Amazon Cloud Technology continued to innovate based on this three-layer architecture, greatly lowering the threshold for the construction and application of generative AI

The newly released Amazon Q is a product that can be customized according to the customer's business. It can meet the needs of various office scenarios and is known as an important tool for generative artificial intelligence application innovation. Amazon Q can be widely used in various vertical industries and will completely change the way industry customers build, deploy and apply generative artificial intelligence on cloud platforms. It can also leverage enterprise private knowledge to complete various tasks, customized according to the unique business, data, code and operations of industry customers. It can also be used in conjunction with other Amazon Cloud Technology products to help enterprises improve productivity and optimize operations. It is understood that Amazon Q has already provided a preview version to customers, Amazon Q in Amazon Connect has also been officially launched, and Amazon Q in Amazon Supply Chain will also be available soon

Amazon Bedrock, which has attracted much attention, has released more model choices and powerful functions to help safely build and scale generative AI applications. The latest high-performance models from Anthropic, Cohere, Meta, Stability AI and Amazon provide customers with richer model selection and new capabilities for evaluating models, simplifying the way to customize models with relevant and proprietary data, and providing the ability to automate complex tasks. Tools that enable customers to build and deploy applications responsibly.

It is particularly worth mentioning that Amazon Cloud Technology has also launched five new Amazon SageMaker functions to make it easier and faster for enterprises to build, train and deploy machine learning models that support various generative AI usage scenarios. Among them, Amazon SageMaker HyperPod can accelerate basic model training on a large scale, shorten training time by 40%, and ensure that the training process that lasts for weeks or months is not interrupted; Amazon SageMaker Inference inference function can reduce deployment costs by 50% and 20% on average. inference latency; Amazon SageMaker Clarify helps customers evaluate, compare, and select the best model; two enhancements to Amazon SageMaker Canvas—preparing data with natural language instructions and leveraging models for large-scale business analysis—enable customers to easily integrate Generative AI integrated into workflow

The implementation of generative AI in industry scenarios has been accelerated

With the continuous upgrading of generative artificial intelligence, innovative application scenarios in various industries are gradually implemented and entering the fast lane. This is a field full of opportunities, but also faces many unknown challenges

According to data released by McKinsey, generative artificial intelligence technology will create approximately US$7 trillion in value for the global economy, while increasing the overall economic benefits of artificial intelligence by approximately 50%. China is expected to contribute approximately US$2 trillion, accounting for nearly 1/3 of the global total

However, although the overall "joy" situation seems good, we cannot ignore the structural "worry". “At present, only the electronics industry has a penetration rate of artificial intelligence exceeding 10% among traditional domestic industries, while the penetration rate in the automobile, petrochemical, pharmaceutical and other industries is between 5% and 10%, and the penetration rate in traditional industries such as building materials is Less than 5%.”

In this case, the field of generative artificial intelligence urgently needs a large number of successful actual cases to produce a significant demonstration effect and provide reference for explorers from all walks of life. Amazon Cloud Technology has accumulated rich practical experience in industries such as automobile manufacturing, life sciences, retail e-commerce, games, and financial services, providing guidance for the application of generative artificial intelligence in actual scenarios

Take the automotive and manufacturing industries as an example: Amazon IoT SiteWise Edge preview is a native software that easily collects, organizes, processes, and monitors device data to help simplify, accelerate, and reduce the cost of sending industrial device data to Amazon Cloud Technology cost; the preview version of vision system data from Amazon IoT FleetWise allows car companies to efficiently collect vehicle data and manage it effectively; the Amazon EC2 DL2q instance launched based on Qualcomm AI 100 helps OEM manufacturers accelerate the development of autonomous driving functions.

At the 2023 Amazon Cloud Technology re:Invent conference, many customers in the automotive and manufacturing industries used Amazon Cloud Technology solutions to carry out application innovation around the two key links of customer journey and product journey. For example, BMW and Honda respectively relied on Amazon Cloud Technology to build next-generation autonomous driving platforms and realize software-defined mobility; BYD used Amazon Cloud Technology to deploy intelligent network connection platforms and Amazon Music and other services, improving the efficiency of automobile research and development and improving the efficiency of automobile research and development. Improved the in-car experience; SAIC chose Amazon Cloud Technology for overseas travel to build an intelligent network solution for its overseas independent brand cars

Life science is also a stage for generative AI to show its talents. Amazon Cloud Technology launches AI recommendations for descriptions in Amazon DataZone to help life sciences customers improve data discovery, data understanding, and data usage by enriching business data catalogs; NVIDIA introduces DGX Cloud and BioNeMo to Amazon Cloud Technology to enable pharmaceutical companies to use data to simplify and accelerate models Training drives drug discovery; Amazon HealthScribe is a HIPAA-compliant generative AI service that assists medical application builders in automatically creating preliminary clinical documents from conversations between patients and clinicians.

Judging from specific implementation cases, both giants and start-ups in the biomedical field have been beneficiaries of generative AI. For example: Based on the migration of applications, databases and servers to the cloud, Amazon Cloud Technology helps Pfizer save more than 47 million US dollars per year, increase data generation speed by 75%, and has achieved innovative breakthroughs in 17 use cases; Amgen uses Amazon HealthOmics to integrate genomics Data is transformed into insights to provide drug treatments for patients; Gilead uses generative AI to accelerate the evaluation of potential targets and promote drug discovery.

Work together to outline the future of enterprise-level generative AI

It is worth noting that the application of generative artificial intelligence in retail e-commerce, games, finance and other industries is accelerating. More and more companies have found effective business transformation and application innovation paths. Enterprise-level generative artificial intelligence Intelligence has reached a critical moment when it explodes

In this critical period, it is impossible to achieve the expected goals by relying on a single breakthrough. We urgently need to build an enterprise-level generative artificial intelligence ecosystem. According to a survey by IDC Consulting, more than 30% of enterprises regard public cloud platforms as the most important strategic partners for generative artificial intelligence. This is the source of change

can be rewritten like this: It can be seen that cloud service providers play a core role in the entire ecosystem. The strategic choices and action paths made by Amazon Cloud Technology at critical moments have established a good foundation for the development of enterprise-level generative AI. The new standards will also attract more participants to join the ranks

From a longer-term perspective, the process of generalization of artificial intelligence has just begun, and digital and intelligent upgrades in all walks of life are still on the way. Enterprise-level generative artificial intelligence is more like a vast wilderness, and it is unknown how many roads lead to the "oasis". Let’s ride horses and whip and meet at the next milestone

The above is the detailed content of Generative AI is accelerating its implementation: Industry application innovation is ushering in the 'cloud moment'. For more information, please follow other related articles on the PHP Chinese website!