What’s New in Artificial Intelligence from the 2022 Gartner Hype Cycle

Hype Cycle for Artificial Intelligence, 2024

The Gartner Hype Cycle™ for Artificial Intelligence (AI), 2022 identifies must-know innovations in AI technology that go beyond everyday AI to add intelligence to previously static business applications, devices and productivity tools.

“It is noteworthy that the Hype Cycle for AI is filled with innovations that are expected to deliver large and even transformative benefits,” said Afraz Jaffri, principal analyst at Gartner. “Particular attention is paid to innovations that are expected to achieve mainstream adoption within two to five years, including composite AI, decision intelligence and edge AI. Early adoption of these innovations can deliver significant competitive advantage and business value and mitigate issues associated with AI model brittleness.”

Download Now: A Detailed Guide to Gartner’s Top 10 Strategic Technology Trends for 2023

AI Innovations in Four Categories

The broad range of AI innovations is expected to impact people and processes inside and outside the enterprise, making it important for many stakeholders, from business leaders to enterprise engineering teams responsible for deploying and operating AI systems, to understand these innovations.

However, data and analytics (D&A) leaders can achieve the greatest benefits by using the Hype Cycle outlook to develop an AI strategy for the future and using technologies that are having a significant impact today.

AI innovations on the Hype Cycle reflect complementary and sometimes conflicting priorities in four main categories:

Data-centric AI

Model-centric AI

Application-centric AI

Human-centric AI

Hype Cycle for AI 2022

Learn more: The Executive Guide to AI

Data-centric AI

The AI community has traditionally focused on improving the results of AI solutions by tuning the AI models themselves, but data-centric AI shifts the focus to enhancing and enriching the data used to train algorithms.

Data-centric AI disrupts traditional data management in addressing AI-specific data considerations, but organizations investing in AI at scale will evolve to retain the evergreen classic data management ideas and extend them to AI in two ways:

Add capabilities needed to facilitate AI development for an AI-focused audience that is not familiar with data management.

Use AI to improve and enhance the classics of data governance, persistence, integration, and data quality.

Innovations in data-centric AI include synthetic data, knowledge graphs, data labeling, and annotation.

Download now: A framework for capturing the business value of AI

For example, synthetic data is data that is artificially generated rather than obtained through direct observation of the real world. Data can be generated using different methods, such as statistically rigorous sampling from real data, semantic methods, and generative adversarial networks, or by creating simulated scenarios where models and processes interact to create entirely new event datasets.

The adoption of synthetic data is increasing across industries with computer vision and natural language application areas, but Gartner predicts that the use of synthetic data will increase significantly as:

Avoid the use of personally identifiable information when training machine learning (ML) models with synthetic variations of original data or synthetic replacement of parts of the data

Reduce ML development costs and save time because it is cheaper and faster to acquire

Improve machine learning performance because more training data leads to better training results

Model-centric AI

While AI approaches have shifted to being data-centric, attention must still be paid to its models to ensure that their outputs continue to help us take smarter actions. Innovations in this area include physics-based AI, composite AI, causal AI, generative AI, foundational models, and deep learning.

Composite AI refers to the fusion of different AI technologies to improve learning efficiency and broaden the level of knowledge representation. Since no single AI technology is omnipotent, composite AI will eventually provide a platform to solve a wider range of business problems in a more efficient way.

Composite AI is expected to achieve mainstream application in two to five years, and its business benefits may be transformative, empowering all walks of life, giving rise to new business models, and ultimately leading to a major shift in the industry landscape. For example, composite AI:

Brings the power of AI to a wider group of organizations that do not have access to large amounts of historical or labeled data but have rich human expertise

Helps expand the scope and quality of AI applications (i.e., more types of reasoning challenges can be embedded)

Causal AI includes a variety of technologies, such as causal graphs and simulations, which help reveal causal relationships to improve decision-making. Although it will take 5 to 10 years for causal AI to achieve mainstream application, its business benefits are expected to be very significant - it will provide new ways to perform horizontal or vertical processes, significantly increasing the company's revenue or saving costs. The benefits of causal AI include:

Increasing the efficiency of causal AI models with smaller datasets by adding domain knowledge

Decision enhancement and autonomy of AI systems

Improving explainability by capturing easily interpretable causal relationships

Enabling greater robustness and adaptability by leveraging causal relationships that remain valid in changing environments

Reducing bias in AI systems by making causal relationships more explicit

Application-centric AI

Innovations here include AI engineering, decision intelligence, operational AI systems, ModelOps, AI cloud services, smart robots, natural language processing (NLP), self-driving cars, smart applications, and computer vision.

Decision intelligence and edge AI are expected to achieve mainstream adoption within two to five years with transformative business benefits.

Decision intelligence is a practical discipline that improves decision making by explicitly understanding and designing how decisions are made and how to evaluate, manage, and improve outcomes through feedback.

Decision Intelligence helps:

Reduce technical debt and improve visibility, and increase the impact of business processes by significantly enhancing the sustainability of organizational decision models based on relevance and transparency quality, making decisions more transparent and auditable

Reduce the unpredictability of decision outcomes by properly capturing and accounting for uncertainty in the business environment and making decision models more resilient

Download now: How to Strategize Your Decision Process

Edge AI refers to the embedding of AI technologies in Internet of Things (IoT) endpoints, gateways, and edge servers for applications ranging from self-driving cars to streaming analytics. Its business benefits include:

Improve operational efficiency, such as manufacturing visual inspection systems

Enhance customer experience

Reduce decision latency with local analytics

Reduce connectivity costs and reduce data traffic between edge and cloud

Solutions are always available regardless of network connectivity

Human-centered AI

This group of innovations includes AI Trust, Risk, and Security Management (TRiSM), Responsible AI, Digital Ethics, and AI Maker and Education Kits.

When AI replaces human decision making, it amplifies both good and bad outcomes. Responsible AI addresses the tension between delivering value and taking risks to achieve the right outcomes. Responsible AI is an umbrella term that encompasses all aspects of making appropriate business and ethical choices when adopting AI, including business and societal value, risk, trust, transparency, fairness, bias mitigation, explainability, accountability, safety, privacy, and regulatory compliance. Responsible AI will take five to ten years to achieve mainstream adoption, but will ultimately have a transformative impact on business.

Digital ethics is a near-term trend (two to five years) that could have a significant impact on business. Digital ethics encompasses a system of values and ethical principles for electronic interactions between people, organizations, and things. These issues, especially those related to privacy and bias, remain a focus for many. People are increasingly aware of the value of their information and are frustrated by the lack of transparency, misuse of information, and information leaks. Organizations are taking action to reduce risks in the management and protection of personal data, and governments are implementing stricter legislation.