Artificial intelligence

Artificial intelligence (AI) is a sophisticated technology that enables computers to emulate human cognitive processes, facilitating everyday tasks and problem-solving. As AI continues to advance, it is becoming an indispensable part of our lives by enhancing the efficiency of our routines both at home and in the workplace.

What is artificial intelligence?

Artificial intelligence definition: Artificial intelligence is the theory and development of computer systems that can perform tasks normally requiring human intelligence. These systems include visual perception, speech recognition, decision-making and translation between languages.

Artificial intelligence includes many theories and technologies and works by analyzing vast amounts of data quickly and efficiently. As it analyzes this data, it can identify patterns that help it carry out tasks that would otherwise be conducted by humans. These tasks can be anything from having a conversation to identifying customers' purchasing patterns.

Recently, with the expansion of resources and research into artificial intelligence, various branches of AI have emerged. Generative AI, machine learning and deep learning represent significant advancements within the field. AI applications are growing and evolving rapidly, promising to extend far beyond our current understanding.

How does artificial intelligence work?

Artificial intelligence is a complex technology, with various subsets and types that operate in different ways. Despite these variations, AI models are generally trained on vast amounts of data before being used to complete tasks based on the learned data.

Here are key steps in the operation of artificial intelligence models:

• Data collection: Gathering diverse data types, including text, numbers, audio, or graphics, depending on the specific AI model
• Data processing: Analyzing data to understand its content and interpret relationships between individual data points
• Learning: Identifying and processing patterns within the data to get a comprehensive understanding of how the training data works and what individual datapoints signifies
• Task performance: Using the data to perform tasks and generate outputs
• Continuous improvements: Learning from their own outputs and becoming more accurate and valuable over time as they are retrained on their successes and failures

This iterative process allows AI models to improve and adapt, making AI a powerful tool for a wide range of applications.

What is the history of artificial intelligence?

Artificial intelligence can be traced back to the pivotal moment when Alan Turing published his seminal paper, “Computing Machinery and Intelligence.” In this work, Turing explored the concept of machine intelligence, laying the groundwork for the development of AI as we know it today.

• 1950s, origins: Soon after this paper was published, computer scientist John McCarthy used the term “artificial intelligence” for the first time. Since this mention in the mid-1950s, there has been a huge shift in interest in the topic and in what it means. 
• 1960s and 70s, initial research: Further research took place in the decades that followed. The first programming language , called list programming, was created, and is still used today. List programming is the foundational language for AI research and development. Some funding and development followed this period of creation. However, the U.S. government showed little to no interest in funding the development of AI research.
• 1980s, increase in global interest and investment: The biggest and most positive shift in artificial intelligence development happened in the 1980s. In this short span of time, the Japanese government allocated approximately $850 million to the Fifth Generation Computer Systems (FGCS) project to develop advanced AI technologies. This significant investment marked a major push in AI research and development. Additionally, the first driverless car was created in Munich, and the first AI-related conference was held at Stanford University.
• 1990s, commercialization and setbacks: The 1990s saw a mix of commercialization and setbacks for AI. While expert systems began to be used in industries such as finance and healthcare, the limitations of AI technologies led to a period known as the “AI winter,” where interest and funding in AI research significantly declined.
• 2000-2010, revival and growth: The early 2000s marked a revival in AI research, driven by advancements in computational power and the availability of large datasets. Machine learning, particularly through the development of algorithms like support vector machines and neural networks , gained prominence. This period also saw the rise of AI applications in areas such as natural language processing and computer vision.
• 2010s, deep learning and ubiquity: From 2010 onward, AI has experienced exponential growth, due to breakthroughs in deep learning. Technologies such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs) have enabled significant advancements in image and speech recognition. AI has become ubiquitous, with applications ranging from semiautonomous vehicles to personalized recommendations and virtual assistants.
• Most recent developments in intelligence: Over the past decade, AI has been integrated into various facets of society and mainstream online media, leading to remarkably innovative developments in a brief period. AI now plays a crucial role in social media algorithms, with models like ChatGPT significantly increasing the volume and quality of online written content. Additionally, AI-generated images are becoming more prevalent online. In some case s, its use presents bad actors with the opportunity to mislead readers. 

In the last five years, AI has seen substantial progress across all domains. Notable advancements in speech recognition and computer vision, coupled with rapid improvements in natural language processing, have significantly accelerated AI’s development trajectory.

• These recent developments are evident in numerous aspects of daily life, including healthcare, where disease detection has become easier, mobility has increased with driver aids and games have become even more sophisticated and visually pleasing. With its prominence in society now, we can only assume it will influence how we live our lives in the future.

Decades of research and interest in artificial intelligence have culminated in the sophisticated technologies we see today. AI is becoming an integral part of people's day-to-day lives, both at home and in the workplace — a far cry from the distant dream it once was.

What are key types of artificial intelligence?

Three key types within the broader domain of artificial intelligence are contributing to the development of AI today.

• Machine learning involves training algorithms to make predictions based on data. There are numerous types of machine learning techniques, each dedicated to various kinds of problem-solving. One of the most common machine learning algorithms is neural networks, named after how the human brain functions. Neural networks are widely used because they can efficiently process and analyze complex data.
• A subset of machine learning, deep learning closely mimics the human brain by enhancing the capabilities of neural networks. Deep learning methods are typically used for tasks that require humanlike intelligence , such as transcribing sound to text.

For example, Zoom uses deep learning to transcribe meetings, allowing participants to focus on speakers and ideas rather than having to take additional notes.

• Generative AI is a subset of AI that can produce entirely new content based on patterns and structures it has learned from existing data. It relies on sophisticated models such as transformers and generative adversarial networks (GANs) to understand and create content. Transformers enable large language models (LLMs) and are more efficient and faster to train.

Transformers have allowed researchers to train larger models without the need for initially labeled data. This capability helps models analyze much more text and information than previously possible.

Advancements in LLMs have paved the way for AI models to generate realistic text and create lifelike images. The development of generative AI models has been rapid in recent years and is expected to continue growing at a similar pace.

How is artificial intelligence used?

At Micron, we not only supply critical generative AI and LLM memory and storage solutions, but we also use AI in our own smart manufacturing processes.

An extremely complex process, silicon manufacturing takes months and involves some 1,500 steps . Micron employs sophisticated AI in every step of this process, dramatically improving accuracy and productivity. Smart sensors are used extensively throughout the manufacturing process to monitor quality and collect complex data in real time. This data becomes input into generative AI workloads that transform operations and achieve historic levels of output, yield and quality for our industry-leading AI memory and storage solutions.

The benefits of Micron’s use of AI extend beyond higher output and yields improved quality and efficiencies, a safer working environment and a sustainable business. The applications are almost limitless. In fact, many big businesses have recognized these applications and are also putting generative AI to use. We know AI because we use AI in our daily operations.