What is artificial intelligence?
Artificial intelligence (AI) is technology that enables machines to demonstrate human-like reasoning and capabilities such as autonomous decision making. Through the assimilation of vast amounts of training data, AI learns to recognize speech, spot patterns and trends, proactively solve problems, and predict future conditions and occurrences.
Artificial intelligence overview
Artificial intelligence is one of the most transformative technologies in modern times. It's also one of the most fast-paced technology disruptions ever. But what is AI, really—and what does it do for business?
The term artificial intelligence originated in 1956 at a scientific conference at Dartmouth College. One of AI’s founding fathers, Marvin Minsky, described it as “the science of making machines do things that would require intelligence if done by men.”
While the core of that definition holds true today, modern AI systems have evolved to demonstrate problem-solving capabilities for tasks like visual perception, speech recognition, planning, decision making, and translation between languages. They can process terabytes of data and insights in real time, proving themselves to be agile, responsive technologies that augment the abilities of human users and increase efficiency, productivity, and satisfaction in the workplace.
Types of artificial intelligence
An AI system is not a single technology but rather an ensemble of technologies that can be combined to perform different types of tasks. Those tasks might be very specific, such as understanding what language is being spoken and responding appropriately, or very broad, such as helping someone with travel suggestions for planning a vacation. But understanding all the different types of technologies that make up AI can be a daunting task. Here are the basics.
Three main types of AI
At the core level, there are three categories of AI:
Narrow AI (also known as weak AI): An AI system designed to perform a specific task or a set of tasks. This is the type of AI used in current applications. It’s called weak not because it lacks power or capability, but because it’s a long way from having the human understanding or consciousness that we correlate to true intelligence. These systems are limited in their scope and do not have the ability to perform tasks outside their specific domain. Examples of narrow AI include voice assistants, facial and speech recognition, and self-driving cars.
General AI (also known as strong AI): In theory, an AI system that would be able to successfully perform any intellectual task that a human could—possibly even better than a human could. Like narrow AI systems, general AI systems would be able to learn from experience and spot and predict patterns, but they would have the capacity to take things a step further, extrapolating that knowledge across a wide range of tasks and situations that are not addressed by previously acquired data or existing algorithms. General AI does not yet exist, though there is ongoing research and development in the field with some promising progress.
Superintelligent AI: An AI system defined as being fully self-aware and surpassing the intelligence of humans. Theoretically, these systems would have the ability to improve themselves and make decisions with above-human-level intelligence. Beyond simply mimicking or identifying human behavior, superintelligent AI would grasp it at a fundamental level. Empowered with these human traits—and further augmented with massive processing and analytical power—it could far exceed our own abilities. If a superintelligent AI system were developed it could change the course of human history, but currently it exists only in science fiction, and there is no known method of achieving this level of AI.
How does artificial intelligence work?
Beyond the major classifications of narrow, general, and superintelligent AI, there are several more different and interrelated levels of artificial intelligence.
Machine learning (ML) is a subset of AI that enables computer systems to learn and improve from experience or data, and incorporates elements from fields like computer science, statistics, psychology, neuroscience, and economics. By applying algorithms to different types of learning methods and analysis techniques, ML can automatically learn and improve from data and experience without being explicitly programmed to do so. For businesses, machine learning can be used to predict outcomes based on analysis of large, complex data sets.
Neural networks are a fundamental component of artificial intelligence, inspired by the structure and function of the human brain. These multi-layered computational models have nodes clustered together like the neurons in a biological brain. Each artificial neuron takes input, performs mathematical operations on it, and produces an output that is then passed to subsequent layers of neurons via fast, parallel processing. During training, neural networks adjust the strength of connections between neurons based on examples in the data, allowing them to recognize patterns, make predictions, and solve problems. They employ a variety of methods to learn from data depending on the task and the type of data. Neural networks have found applications in various fields such as image and speech recognition, natural language processing, modeling, autonomous vehicles, and more.
Deep learning (DL) is a data-centric subset of machine learning that uses neural networks with multiple (deep) layers to learn and extract features from vast amounts of data. These deep neural networks can automatically discover intricate patterns and relationships in the data that might not be immediately obvious to humans, allowing for more accurate predictions and decisions. Deep learning excels at tasks such as image and speech recognition, natural language processing, and data analysis. By leveraging the hierarchical structure of deep neural networks, deep learning has revolutionized many domains, including healthcare, finance, and autonomous systems.
Generative AI (gen AI) is a type of deep learning that uses foundation models such as large language models (LLMs) to create brand-new content—including images, text, sound, videos, and software code—based on their training data. Gen AI is a catch-all term for various foundation model technologies—neural networks trained on massive volumes of data using self-supervised learning, such as predicting the next word in text. Its emergent capabilities make it a breakthrough in AI, with a single model sometimes able to write both poems and business documents, create images, and pass reasoning tests. Imagine the output of two LLMs, one trained exclusively on scientific research journals and another trained on sci-fi novels. They may both generate a brief description of the movement of objects in space, but the descriptions would be drastically different. Generative AI has many business applications, such as creating realistic product prototypes, conducting natural conversations in customer service, designing personalized marketing materials, automating content-creation processes, and creating graphics and special effects. Both businesses and consumers have been adopting generative AI at a remarkable rate, driven by the fact that many gen AI applications don’t require programming or coding skills to use—users simply describe what they want using regular language, and the application performs the task, often with impressive results. According to a McKinsey report, in 2023:
33% of organizations are using gen AI regularly in at least one business function.
40% of organizations will increase investment in AI because of gen AI.
60% of organizations employing AI are already using gen AI.
AI applications
Here are a few other ways AI is changing the way people work, learn, and interact with technology:
Robotics
Robotics has been used in manufacturing for years, but before the introduction of AI, calibration and reprogramming had to be done manually—and typically only after something broke down. By using AI—often in the form of Internet of Things (IoT) sensors—manufacturers have been able to greatly expand the scope, volume, and type of tasks their robots can perform, while also improving their accuracy and reducing downtime. Some common examples of AI-assisted robotics include order-picking robots in warehouses and agricultural robots that water crops at optimum times.
Computer vision
Computer vision is how computers “see” and understand the contents of digital images and videos. Computer vision applications use sensors and learning algorithms to extract complex contextual information that can then be used to automate or inform other processes. It can also extrapolate on the data it sees for predictive purposes, such as in the case of self-driving cars.
Natural language processing (NLP)
Natural language processing systems recognize and understand written or spoken language. In more sophisticated applications, NLP can use context to infer attitude, mood, and other subjective qualities to most accurately interpret meaning. Practical applications of NLP include chatbots, call center interaction analysis, and digital voice assistants such as Siri and Alexa.
Learn more about AI
Discover the quick value artificial intelligence can bring to your business with a comprehensive collection of AI-specific resources.
Benefits of AI
AI technologies have moved beyond the early adopter stage and are now mainstream in many business applications.
Today, companies derive measurable benefits from building AI into their core business processes:
Enhanced efficiency and productivity: One of the most significant advantages of AI in enterprise is its ability to automate tasks and streamline operations. AI-powered systems can process large volumes of data at lightning speed, freeing up valuable human resources to focus on more value-added activities. This increased efficiency leads to improved productivity, as employees can devote their time to strategic decision making and innovation rather than routine and mundane tasks.
Improved customer experience: AI technology has revolutionized the way businesses interact with customers. Through NLP and ML algorithms, AI-powered chatbots and virtual assistants can provide personalized and real-time support to customers, 24/7. This availability not only enhances customer satisfaction, but also helps businesses deliver a seamless customer experience across channels while reducing response times and human error.
Data-driven decision making: Enterprise AI systems can analyze vast amounts of structured and unstructured data, enabling organizations to make more informed decisions. Deriving meaningful insights from this data empowers companies to identify trends, predict customer behavior, and optimize their operations. AI algorithms can detect patterns that humans may overlook, providing valuable information for strategic planning, risk assessment, and streamlining business processes.
Operational efficiency: AI can automate repetitive, time-consuming tasks and workflows, as well as handle complex calculations, data analysis, and other tedious tasks with precision, leading to improved accuracy and reduced errors. AI can also help detect anomalies, fraud, and security breaches swiftly, mitigating potential losses.
Enhanced workforce collaboration: AI can foster greater collaboration and knowledge sharing among employees. Intelligent systems can assist in data discovery by providing easier access to relevant information and providing insights that aid employees in making informed decisions. Additionally, AI-powered collaboration tools enable seamless communication and knowledge sharing across teams, departments, and even geographically dispersed locations, encouraging innovation and enhancing productivity.
Enterprise AI in action
The scope and accessibility of modern enterprise AI makes it useful for many fields.
A few examples of AI use cases across industries include:
AI in healthcare: Medical data sets are some of the largest and most complex in the world. A major focus of AI in healthcare is leveraging that data to find relationships between diagnosis, treatment protocols, and patient outcomes. Additionally, hospitals are turning to AI solutions to support operational initiatives, such as workforce satisfaction and optimization, patient satisfaction, and cost reduction.
AI in banking: The financial services industry has been one of the earliest to adopt AI at scale, specifically to accelerate the speed of transactions, customer service, and security response. Common applications include AI bots, digital payment advisers, and fraud detection.
AI in manufacturing: Today’s smart factory is a network of machines, IoT sensors, and computing power—an interconnected system that uses AI and machine learning to analyze data and learn as it goes in real time. AI continually optimizes and informs the automated processes and intelligent systems within a smart factory, from monitoring equipment conditions to forecasting supply chain issues to enabling predictive manufacturing.
AI in retail: Online shoppers are engaging across a wide range of touchpoints and generating larger amounts of complex and unstructured data sets than ever before. To understand and make use of this data, retailers are using AI solutions to process and analyze disparate data sets, improving marketing and providing better shopping experiences.
AI ethics and challenges
While AI presents extraordinary opportunities, it also comes with risks that need to be recognized and mitigated to prevent harm to individuals, groups, businesses, and humanity as a whole. Here are some of the most urgent AI ethics challenges that consumers, businesses, and governments alike should be mindful of as they strive to use AI responsibly.
Ethical use of customer data: By 2029, there will be an estimated 6.4 billion smartphone users worldwide. Each device can share enormous amounts of data, from GPS location to users’ personal details and preferences, as well as social media and search behaviors. As businesses gain wider access to their customers’ personal information, it becomes increasingly important that they establish benchmarks and ever-developing protocols to protect privacy and minimize risk.
AI bias: AI systems can reflect or amplify existing biases present in their training data, potentially leading to unfair outcomes in applications like job hiring or loan approvals. To mitigate these biases, organizations must make sure their datasets are diverse, conduct regular audits, and employ bias-mitigation algorithms. A real-life example of AI bias occurred in the U.S. healthcare system, where an AI model lacking critical bias-mitigation capabilities inferred from training data that demographic groups that spend less on healthcare don’t need as much care in the future as higher-spending groups, resulting in a bias that affected health decisions for hundreds of millions of patients.
AI transparency and explainable AI: AI transparency refers to the openness and clarity of how AI systems work to ensure that their operations, decision-making processes, and outcomes are understandable and interpretable by humans. This is crucial for building trust in AI applications and addressing concerns about bias, accountability, and fairness. Explainable AI specifically focuses on developing AI models and algorithms that can provide explanations for their decisions and predictions in a way that is understandable to users and stakeholders. Explainable AI techniques aim to demystify complex AI systems by revealing the factors and features that influence their outputs—enabling users to trust, verify, and potentially correct AI decisions where necessary.
Deepfakes: The term deepfake is a combination of deep learning and fake. A deepfake is a sophisticated method of creating or altering media content, such as images, videos, or audio recordings, using AI. Deepfakes enable the manipulation of facial expressions, gestures, and speech in videos, often in a remarkably realistic manner. This technology has garnered attention due to its potential for creating convincing but fabricated content that can be used for various purposes, from entertainment and artistic expression to more concerning applications like misinformation and identity fraud.
Learn more about artificial intelligence
Explore AI built for real-world results
See how you can benefit from AI that’s built into your core business applications, connecting your people, data, and processes.
Meet Joule—the AI copilot that truly understands your business
Revolutionize how you interact with your SAP business systems, making every task simpler and every touchpoint count.