Session
WebAI - Exploring agentic capabilities within the browser
Thursday 23 July
15:00 – 16:00
Educator 2
Back to agenda
AI carries a lot of hype and it's most often known via the lens of a "chatbox", however there are many more ways where it's being adopted, and one such place is natively inside the browser. There have been significant development in this area by browser vendors and in this session we will explore these proposed standards and how to use them today in all browsers.
We are talking here about the
- Prompt API: Sends prompts to built-in Gemini Nano for text generation.
- Summarizer API: Summarizes text content.
- Translator API: Translates text on-device without server-side processing.
- Language Detector API: Identifies the language of text.
- Writer API / Rewriter API: Assists with content creation, rewriting, and refining text.
- Proofreader API: Provides grammatical and punctuation corrections.
All of these are able to work in-browser, offline, or optionally be polyfilled and use another agent like Gemini or OpenAI models as the backend.
There's a lot of work being done to starndardized these by the W3C
The goal is to demonstrate an alternative approach to the ai world, which can be private and offline.
We are talking here about the
- Prompt API: Sends prompts to built-in Gemini Nano for text generation.
- Summarizer API: Summarizes text content.
- Translator API: Translates text on-device without server-side processing.
- Language Detector API: Identifies the language of text.
- Writer API / Rewriter API: Assists with content creation, rewriting, and refining text.
- Proofreader API: Provides grammatical and punctuation corrections.
All of these are able to work in-browser, offline, or optionally be polyfilled and use another agent like Gemini or OpenAI models as the backend.
There's a lot of work being done to starndardized these by the W3C
The goal is to demonstrate an alternative approach to the ai world, which can be private and offline.
