Many of the AI technologies we use today run in massive data centers operated by cloud computing giants such as Amazon, Microsoft, and Google. But with a technology called WebGPU that Google is now building into its Chrome browser, web apps can take advantage of AI processing more directly.
Google announced the adoption of WebGPU. Google I/O Conference Wednesday. With WebGPU, web apps on phones and laptops can better take advantage of the kind of artificial intelligence software that’s sweeping everything from creative tools to health apps.
“WebGPU makes the web AI-enabled,” Matt Waddell, who leads Chrome’s developer and consumer focused work, said in an exclusive interview ahead of the conference. He said the company will be demonstrating at Google I/O a web app running Stability AI’s Stable Diffusion software for converting text prompts into images.
The move reflects the widespread adoption of AI technologies that have been used behind the scenes for years, but with the emergence of new generative AI tools such as OpenAI’s ChatGPT, Microsoft’s Bing, Google’s Bard, and Adobe’s Firefly. , it is becoming more pronounced. Google is desperate to ride on the excitement after being unfairly caught in ChatGPT, even though CEO Sundar Pichai declared Google to be an AI-first company in 2016.
Hardware in phones and laptops is much less powerful than server hardware in data centers, but AI can run locally on the device, avoiding network issues and helping control data. This could be useful for health app companies that want to keep sensitive data and results private, Waddell said.
Chrome Consumer and Developer Project Lead Matt Waddell said at Google I/O that the WebGPU interface will accelerate AI tasks on the web.
Stephen Shanlkand/CNET
WebGPU’s origins can be traced back to projects by Google, Apple, and others that revolutionized video game hardware on the web. This allows web apps to take advantage of the raw power of the graphics processing unit (GPU), previously only available to graphics-intensive video games. The computing industry has discovered that these GPUs are also very good at accelerating AI.
AI software can also run in apps that run natively on the device, just like Adobe’s Lightroom and Photoshop image editing apps. But his faster AI for web apps means more opportunities for developers to leverage the versatility of his web platform.
“We are always thinking about a more functional and powerful web platform,” said Waddell.
Reduces browser compatibility issues
But the ubiquity of the Web also comes with problems, such as piecemeal efforts to add new browser features that programmers can use. This undermines the ease of developing websites and web apps that work the same on many devices without complex workarounds.
But even at Google I/O, the company announced partnerships with Mozilla’s Firefox, Apple’s Safari, and Microsoft’s Edge, and a new initiative called Baseline that flags web features that can be relied upon by any browser that developers support. Waddell said he detailed
Baseline 2024 is expected to arrive at the end of this year, after which it will be continuously updated as new compatible features arrive.
Move your Android apps to the web faster
Another programming change unveiled at Google I/O is an enhanced web technology called WebAssembly, or Wasm for short, that promises to speed up web apps. Wasm allows a programmer to convert native software to run inside a web browser, powering his web versions of Adobe Photoshop and Autodesk’s AutoCAD design software.
The company will announce at Google I/O that developers can now convert apps written in the Kotlin language to Wasm. Kotlin is the most widely used language for creating Android apps, so it makes it easier for Android developers to reach new users of other platforms.
Editor’s Note: CNET uses an AI engine to create personalized financial explanations that are edited and fact-checked by our editors. For more information, see this mailbox.
