Everything you need to know about AI can be found here—why AI is important, how AI works with 5G, and the differences between edge computing and distributed intelligence. And of course, how AI improves smartphones. Discover how Qualcomm Technologies is making AI ubiquitous.
A: Artificial intelligence, or AI, is an umbrella term representing a range of techniques that allow machines to mimic or exceed human intelligence.
When humans think, they sense what’s happening in their environment, realize what those inputs mean, make a decision based on them, and then act. Artificially intelligent devices are in the early stages of beginning to replicate these same behaviors.
AI explained in 101 seconds
Sep 27, 2017
AI is the superset of various techniques that allow machines to be artificially intelligent. For an analogy, think of a Russian nesting doll: machine learning is a subset of AI, and deep learning is a subset of machine learning
Machine learning refers to a machine’s ability to think without being externally programmed. While devices have traditionally been programmed with a set of rules for how to act, machine learning enables devices to learn directly from the data itself and become more intelligent over time as more data is collected.
Deep learning is a machine learning technique that uses multiple neural network layers to progressively extract higher level features from the raw input data. For example, in image processing, lower layers of the neural network may identify edges, while higher layers may identify the concepts relevant to a human such as letters or faces.
An AI assistant is a program powered by machine learning that can respond to you, provide information, anticipate your needs, and perform tasks at your request. While these assistants are most commonly thought of in terms of smartphones and smart home speakers, they can exist in a range of devices and will become common in XR glasses, home appliances, connected cars, and more. With the 5th generation Qualcomm® AI Engine on Qualcomm Snapdragon 865, we’re enabling AI assistants with advanced capabilities to enhance user experiences while meeting the power and thermal constraints of mobile devices.
A: AI is a powerful tool for addressing a variety of challenges, such as voice translation or wireless channel estimation, that are difficult to model or solve with traditional methods. For the end user, AI seamlessly offers enhanced experiences, personal assistance, and automation of repetitive tasks. In addition, AI can make devices more energy efficient and allow us to interact with them in more convenient ways, like with an always-on voice user interface.
A: The biggest challenges AI models face are how to be more power efficient, how to learn from less data, how to learn from unlabeled data (unsupervised learning), and how to generalize across multiple tasks. The industry is also focused on making AI unbiased and explainable, such that we know how it works, where it fails, and how to quantify confidence levels. For example, we want to understand how the AI used for autonomous driving is deciding how to drive safely on a road under various environments and weather conditions.
A: AI is currently benefiting from more data and more efficient hardware, as well as better AI tools and networks/algorithms. Advancements in state-of-the-art accuracy for various tasks happen regularly due to the collaborative nature of the AI research community through papers and workshops. For example, Qualcomm AI Research has published many papers in the areas of power efficiency, personalization, and efficient learning. These advancements are being applied in more areas, integrated into different types of devices, and enhancing our user experiences.
A: When there’s bias in the data set that trains an AI model, the model will contain the same bias. A way to address this bias is by collecting robust and diverse data. For example, in distributed learning on smartphones, you may want to sample data from different geographies and demographics. If bias is noticed in algorithms, the data should be examined to determine whether new data should be added.
A: 5G and AI are two of the most disruptive technologies the world has seen in decades. While each individually is revolutionizing industries and enabling new experiences, their combination is going to be truly transformative. The low latency and high capacity of 5G can allow AI processing to be distributed among the device, edge cloud, and central cloud. This enables flexible system solutions for a variety of new and enhanced experiences like vastly improved voice UI and boundless XR.
Edge computing brings computation and data storage closer to the location where it is needed, saving bandwidth and improving response time as a result. Edge computing may have different meanings depending on context: the edge of a network, the edge of a cloud, or the edge device. For example, edge computing could mean on-device processing for a smartphone or IoT device, or it could mean processing in the edge cloud or close to a base station for a cellular network provider.
A: Edge computing is computing that happens at the edge cloud or edge device, while cloud computing occurs in the central cloud. Where the processing is located may result in different levels of performance, latency, or privacy. Both offer different benefits and are complimentary to each other.
A: Computer vision enables computers to gain high-level understanding from digital images or videos. Deep learning has brought tremendous advancements in computer vision.
A: Computer vision involves generating feature detectors, which traditionally have been hand-crafted by humans. With the help of large data sets of labeled images or videos, machine learning, and specifically deep learning, can learn these feature detectors automatically and more accurately than humans.
A: Distributed intelligence is the result of AI processing that happens jointly on the device, edge cloud, and/or central cloud processing. A low-latency, high-reliability, and high-capacity link is essential for enabling distributed intelligence and allowing workloads to be processed in the most appropriate place. This is why 5G is so important for enabling distributed intelligence for a variety of use cases, such as extended reality (XR) and the factory of the future.
Distributed intelligence is the result of running artificial intelligence algorithms across various computing devices, such as a phone and a server in the cloud.
A: Edge computing has multiple meanings, but it is generally a specific location where the computing occurs. Distributed intelligence means that the intelligence is being created from running the AI algorithms in multiple locations by splitting the processing workloads.
A: AI is running on your phone behind the scenes for a variety of use cases, inferencing neural networks on your device to help you take better photos, understand a different language, identify music, and assist with gaming. This is also known as on-device AI. AI can also be used on mobile phones for real-time language translation, Qualcomm Technologies is one of the firsts in the industry to introduce this feature It also improves user experiences on Snapchat, making lenses smoother and faster. Beyond that, cameras on mobile phones also use AI for low-light photography, super-resolution, best shot, scene settings, switching between camera lenses, and more.
A: We began researching AI more than a decade ago. Today, more and more intelligence is moving to end devices, and mobile is becoming the pervasive AI platform. Building on the smartphone foundation and the scale of mobile, Qualcomm Technologies envisions making AI ubiquitous—expanding beyond mobile and powering other end devices, machines, vehicles, and things. We are inventing, developing, and commercializing power-efficient on-device AI, cloud AI, and 5G to make this a reality.
©2021 Qualcomm Technologies, Inc. and/or its affiliated companies.
References to "Qualcomm" may mean Qualcomm Incorporated, or subsidiaries or business units within the Qualcomm corporate structure, as applicable.
Qualcomm Incorporated includes Qualcomm's licensing business, QTL, and the vast majority of its patent portfolio. Qualcomm Technologies, Inc., a wholly-owned subsidiary of Qualcomm Incorporated, operates, along with its subsidiaries, substantially all of Qualcomm's engineering, research and development functions, and substantially all of its products and services businesses. Qualcomm products referenced on this page are products of Qualcomm Technologies, Inc. and/or its subsidiaries.
Materials that are as of a specific date, including but not limited to press releases, presentations, blog posts and webcasts, may have been superseded by subsequent events or disclosures.
Nothing in these materials is an offer to sell any of the components or devices referenced herein.