Generative AI - From Imagination to Reality
Generative AIs are artificial intelligences that are designed to analyze text, images, video and other content formatted for human consumption, and use this data to create additional content that people can read, view, or listen to. ChatGPT from OpenAI is among the most famous examples of a generative AI – the G in GPT stands for “generative.” Generative AIs can accept a wide variety of natural language inputs (text, speech, even gesture), process those requests, and provide a response in the appropriate format, which could be text, video, or audio. That can include intermodal interactions, turning text to speech, for example. People will increasingly rely on generative AIs to provide information and support.
Generative AIs are typically deep-learning models. Deep learning is a specific type of neural network; neural networks are a specific type of machine learning (ML) that are set up to process information in a manner designed to mimic the way brains function. Generative AIs are almost always trained extensively on existing data (research papers, image libraries, music catalogs, fiction, etc.). Generative AIs can also be instructed to generate new content that previously did not exist – a novel, a movie script, an image, even video.
Some of the most common use cases for generative AIs are:
- Personal Assistants
- Voice Assistants – (Alexa enhancements for smart homes)
- Changing appearance on screens (vision related)
Key Generative AI Applications
Generative AI at the Edge
The Edge GenAI Chatbot (demonstrated at electronica 2024) is a locally operated chatbot that runs directly on embedded devices, delivering fast, low-latency responses while ensuring enhanced privacy. Its modular software design allows flexibility across diverse hardware configurations, making it adaptable to specific application requirements. Additionally, it is compatible with a broad selection of TRIA System-on-Modules (SOMs) to optimize performance. Comprehensive software support is provided, offering robust resources and tools for seamless integration and customization.
Learn MorePersonal Assistants
Personal assistants are designed to respond to queries. Queries can be fairly simple requests to provide a single data point (“where is the nearest gas station?”), but they can also require research on complex subjects, with the requested result to be a report complete with high-level analysis (“provide a synopsis of ‘Gravity’s Rainbow’ and explain the major themes of the book”). Personal assistants of this nature can be installed on potentially any computer-based system; chat assistants (e.g. ChatGPT) and similar are now widely accessible through computers, laptops, and smartphones.

Voice Assistants
Voice assistants are a specific type of personal assistant in that the input is always spoken language. There are any number of well-known examples, among them Alexa (Amazon) and Siri (Apple). The first voice assistants were based on less sophisticated AIs and were capable of responding only to simpler requests (“play ‘Hotel California’ on Spotify”), but they are obvious candidates for generative AI. Voice assistants tend to be bundled into systems where hands-free operation is likely to be preferred. Originally voice assistants were built into smart speakers, but voice assistants are appropriate in other environments, including passenger vehicles, for example.

Modifying Appearance on Screens
AI revolutionizes the way appearances are altered on screens through advanced algorithms that analyze, interpret, and manipulate visual data in real-time. From enhancing facial features to applying virtual makeup or altering backgrounds seamlessly, AI-powered software utilizes deep learning techniques to understand and modify digital content with unprecedented precision. By harnessing machine learning models trained on vast datasets, these technologies can intelligently adapt to diverse scenarios, ensuring natural-looking transformations and immersive experiences for users across various applications, including video conferencing, gaming, and virtual reality. A prime example of this is the Kontron x86 AMD Industrial Platform that was demonstrated on Avnet Silica's booth at Embedded World 2024 and electronica 2024.
Martin Grossen demonstrates the Kontron x86 AMD Industrial Platform at electronica 2024
onsemi Image Sensing Technologies
As a leader of intelligent sensing technology, onsemi develops innovative solutions with customers and specific applications in mind. Leveraging its deep technical expertise, onsemi's sensing products are designed to solve application-focused challenges, delivering efficient and optimized performance. onsemi offers a diverse portfolio of products serving a range of markets, including automotive, industrial, and commercial. Learn more about the:
- AR0820AT
- AR0822
- AR0823AT
Learn More
The Evolution of Generative AI
Generative AI is not new. It has, however, recently taken an enormous leap in what it can do, inspiring no little amazement, as well as a bit of alarm.
A popular candidate as the first generative AI is the first chatbot, ELIZA, introduced in the 1960s, which was trained to respond like a psychologist. The subsequent history of AI research was marked by AI “winters” alternating with periods of resurgence typically coincident with some applicable technological advance, such as the introduction of microprocessors or the formulation of AI techniques such as back-propagation.
Learn MoreTools
Design Hub
Browse and review over a thousand proven reference designs to accelerate your design process. Try our design tool and then export it to your CAD tool of choice.

Training & Events
Learning for better, faster projects builds
Connect with the Avnet Silica experts who will guide you to reach further with your projects with on-going seminars, workshops, trade shows and online training.

Market
Smart City
Smart City is an opportunity to reimagine our cities with the goal to change our quality of life and to supply clean water and sufficient energy for all citizens in an environmentally sustainable and equitable way. Let's make our cities smart.

Contact us
Have a question?
Int. Freecall - 00800 412 412 11 | Product or shop-related inquiries: OnlineSupportEU@avnet.com | Content-related questions: yourmessage@avnet.eu
