STACK #239 September 2024

TECH FEATURE

visit jbhifi.com.au/stack

SMARTER PHONES ARE HERE!

The rise of generative AI and its integration into new smartphones has revolutionised how we use our devices, including easy responses to emails, document summarisation, image and text generation, and more. Let’s take a closer look at this amazing new tech. Words Scott Hocking

What is generative AI? Generative AI differs from other AI models in that it can create new content in real time, such as wallpapers, images, and text. It can also alter backgrounds in photos, concisely summarise key points from lengthy documents, transcribe voice recordings, and eliminate switching between apps when sharing photos or searching online. It also works to boost privacy and security, improves the overall efficiency of your smartphone, and can quickly adapt to your needs. Popular examples of generative AI chatbots include ChatGPT, Microsoft Copilot, and Google Gemini (see page 12), and if you’ve used any of these you’ll appreciate the exciting potential of generative AI integration on a smartphone. How does it work? Generative AI uses machine-learning algorithms to learn your habits, preferences, and how you interact with your phone. It doesn’t understand the data it’s learning, though, it simply recognises patterns and structures in the data and uses them to reproduce or create new content at blazingly fast speeds. The more data the algorithms learn, the faster and more efficient they become in carrying out AI-powered tasks, such as generating new and original content based on a prompt, text, photos, audio, and even a rudimentary sketch. What are the advantages of using generative AI? There are plenty! Think of generative AI as your

as well as across a range of editing tools like Google’s Magic Eraser that scrubs out unwanted objects in photos, and Samsung’s Generative Edit for moving, resizing, and erasing objects within an image. Where it all happens A smartphone’s respective chipset (such as Qualcomm’s Snapdragon 8 Gen 3 processor and Google’s Tensor G4 chip) – or more specifically, its NPU (Neural Processing Unit) – can perform the processing of many AI-powered tasks directly on the device. This means there’s no transmission of data to and from the cloud for processing, and certain tasks don’t require an internet connection. This offline, native AI processing has many advantages: along with faster response times, you can also save on data usage with less back

own personal assistant and content creator. It can offer intelligent suggestions and context aware assistance to accomplish tasks more efficiently and effectively, and free you from mundane time-consuming chores, so you can be more productive and creative elsewhere. For example, generative AI can be an invaluable research tool, saving hours of online searching and note scribbling. It can search for and summarise in a flash the info you need on a certain subject, and deliver complex info in simple terms. Need to add an image to a document but can’t find the right one? Simply draw a basic sketch of what you need and AI will bring it to photorealistic life! Travelling abroad but don’t speak the local language? AI is your personal interpreter. AI is also at work in many phone camera systems to improve low-light photography,

and forth between the device and cloud, and storing data on the device itself creates an extra layer of privacy. Some smartphones use a hybrid model that requires certain AI-powered tasks

to be handled by the cloud, like Google and Samsung’s Circle to Search function, while things like generative photo editing and live language translation can use both cloud and native processing. Running AI functions natively not only boosts speed and productivity, it also enables a more personalised user experience.

*GalaxyAI features will be free until the end of 2025 on supported Galaxy devices, after which you may be charged service fees to use GalaxyAI features provided by a third party.

SEPTEMBER 2024

jbhifi.com.au

8