AI is too complex and dynamic a technology to be approached one-sidedly, only from the business or IT side. There will be a complete merger of AI management processes with data management processes and cloud technologies, which will require a new approach to management providing AI. For such investment in AI to really pay off, it must be built into application systems that can run 24/7. These systems, in turn, require cloud-based computing power that can scale up and down in performance to cost-effectively meet ever-changing needs.
It is also important for researchers to note that leading companies clearly distinguish between different stages of the management life cycle in these three interrelated areas. They are constantly reviewing the strategy against business needs, making adjustments to its implementation, taking into account which data models, ethical AI and computing power can meet these needs. And they are also streamlining operations to improve data collection, models used, and cloud technologies. When data, AI and cloud technologies seamlessly combine at all stages, a flexible and powerful system is created to help you determine exactly what data is needed, collect or synthesize it, and use it to reduce risks and discover new opportunities.
Modeling and simulation will unlock the potential of AI in supply chains, the metaverse and beyond
The idea of modeling and simulating real processes is nothing new, but modeling combined with AI could make it revolutionary, especially for tasks having high capital intensity, unique structure and rare repeatability. Simulation can help business leaders test an infinite number of scenarios to make the right decisions in the short and long term. AI, for example, can create "digital twins": detailed models of physical assets such as aircraft engines, oil rigs, and even entire cities. IoT technologies bring these models to life and make it possible to create digital equivalents of manufacturing enterprises and/or smart cities.
In combination with AI, digital twins can predict the behavior of consumer groups or create digital copies of them. Large-scale AI-based simulations can recreate and predict the potential behavior of financial assets and markets. AI will be a fundamental element in the development of simulation technology to the level of the formation of the "metaverse" - the convergence of technological trends that allow users to perceive our digital world in a new way, rising to a new level of autonomy and freedom.
Consideration should be given to incorporating AI models into strategy, where changing consumer preferences, competitor actions, and regulatory policies can be assessed. By combining multiple AI-generated models, you can also create a more resilient, transparent, and cost-effective supply chain by modeling your supplier behavior, market dynamics, and potential disruptions.
It will be possible to estimate and predict the full cost of AI, not just cost savings
In many cases predicting the return on investment in AI is quite difficult - the technology is complex and constantly evolving. It can be challenging even to determine the value that an already working AI system provides. With the estimation of direct effects, everything is straightforward - the value is provided by the operational use of AI tools, for example, processing invoices or purchase orders (“boring AI”). This directly affects the cost of operations through payroll levers, the time it takes to complete operations, and the amount of scrap when solving typical situations.
However, the value of more advanced ways of using AI is much more problematic to determine, the researchers say and ask clarifying questions. How, for example, to determine the value of a better strategic decision, or how to determine the exact cost of preventing a disruption in the supply chain by collecting social media signals and processing them with AI that issued an early warning? Fortunately, new methods of evaluation can take into account both direct benefits and costs, such as increased productivity or equipment costs, and “indirect” benefits and costs, for example, improved interaction with staff or time costs. profile specialists. Leading companies are beginning to take a portfolio approach to AI investments to increase the likelihood that a successful application will outweigh any potential failure many times over.
AI will be too important to be managed by AI experts
Managing AI is associated with particular challenges that are closer to those arising from managing people rather than computing techniques. AI is a moving target, constantly operating on sensitive data and supporting ever more business-critical decisions and actions. It's also an incredibly complex technology that does things (like doing creative work) that no other technology has ever been able to do. The best solution here, according to the authors of the study, could be a management triad - end-to-end management of the life cycle of the "data - AI - cloud technologies" (DAC) complex from the side of risk assessment, digitalization and business. This corporate governance model provides for new procedures, roles and responsibilities for each of the three lines of defense. Each will play a role in determining whether an AI solution should be included in development, production, or operations. If the decision is incorporated into operations, each line will also help decide when it should be retrained, rebuilt, or decommissioned.