Produced | CSDN Cloud Computing
Contribution | Gartner
Gartner today released a list of key trends shaping the future direction of data science and machine learning (DSML). With the rapid development and evolution of the DSML industry, the importance of data for the development and application of artificial intelligence (AI) is increasing, especially the focus of investment is also shifting to the field of generative artificial intelligence.
Peter Krensky, research director at Gartner, said: "As the application of machine learning in various industries continues to expand rapidly, DSML is also shifting from a purely predictive model to a more pervasive, dynamic and data-centric technology field, and generative artificial intelligence The wave of intelligence (AI) is also fueling this trend. While potential risks continue to emerge, so too do new capabilities and use cases for data scientists and their organizations.”
According to Gartner research, important trends affecting the future direction of the DSML industry include:
Trend 1: Cloud Data Ecosystem
Data ecosystems are transitioning from standalone software or hybrid deployment models to fully cloud-native solutions. Gartner predicts that 50% of new cloud deployments by 2024 will be based on a consistent cloud data ecosystem rather than manually integrated point solutions.
Gartner recommends that organizations evaluate the two capabilities of the data ecosystem: one is to solve the problem of data decentralization; the other is to access and integrate with data outside their own environment.
Trend 2: Edge AI
Organizations increasingly need to create and process data at the edge through edge AI, which will help organizations gain real-time insights, discover new business models, and meet stringent data privacy requirements. Edge AI can also help organizations improve AI development, orchestration, integration, and deployment capabilities.
Gartner predicts that by 2025 more than 55 percent of deep neural network data analysis will occur at the point of data capture at the edge, up from less than 10 percent in 2021. Organizations should determine which applications, AI training, and inference capabilities need to be moved to edge environments near IoT endpoints.
Trend 3: Responsible AI
Responsible AI enables AI to be a positive force rather than a threat to society and AI itself. When organizations need to use AI to make the right choices in terms of business logic and ethics, they will encounter many issues, such as business and social value, risk, reputation, transparency, and accountability. Responsible AI can help address these issues. Gartner predicts that by 2025, 1% of AI service providers will use pre-trained AI models on a large scale, bringing responsible AI into the focus of society.
Gartner recommends that organizations should consider the risk factor when tapping the value of AI, and be cautious when using AI solutions and models. Vendors should be required to provide assurances that they are managing their own risks and compliance obligations to prevent potential financial loss, legal action and reputational damage to organizations.
Trend 4: Data-centric AI
This approach is no longer centered on models and codes, but on data to create a more powerful AI system. Organizations will adopt solutions such as AI-specific data management, synthetic data, and data labeling technologies to address many data challenges such as data accessibility, volume, privacy, security, complexity, and scope.
Creating synthetic data using generative AI is a rapidly developing field, a technique that offloads the burden of obtaining real-world data to more efficiently train machine learning models. Gartner predicts that by 2024, 60% of AI data will be synthetic data, which will be used to simulate reality, future scenarios and reduce AI risks, compared with only 1% in 2021.
Trend 5: Accelerating AI investment
Organizations entering the solution implementation phase, as well as industries looking to grow through AI technology and related businesses, will continue to accelerate investment in AI. Gartner predicts that by the end of 2026, AI startups relying on foundational models (large models trained with massive amounts of data) will receive more than $10 billion in investment.
In a recent Gartner survey of more than 2,500 business executives, 45 percent of respondents said the recent ChatGPT craze has prompted them to increase their investments in AI. Seventy percent of respondents said their organizations were in the research and exploration phase of generative AI, and 19 percent said their organizations were in pilot or production stages.