AI · Official
Google Research Unveils Agile Deliberation for Refining AI Gray-Area Rules
Google Research blog ·
At a glance
Google Research introduced a method called Agile Deliberation. When people teach an AI image classifier a concept with blurry boundaries, such as 'healthy food,' the AI asks questions in a conversational way and helps refine the criteria together. The goal is to help the AI set accurate standards for which photos to select.
Key points
- People often do not have perfect criteria from the start, and they learn the boundaries of a concept by looking at examples.
- The method defines a concept by what fits (included) and what does not (excluded), and it shows photos near the boundary line.
- When people give feedback on boundary-line photos, the AI automatically revises its instructions.
- The researchers ran experiments with 18 real people for 1.5 hours each, and reported that users got less fatigued and the resulting model was better.
Why it matters
The way AI learns people's fuzzy judgment criteria may increasingly be used for tasks without fixed right answers, such as content review and recommendations.
Original source
Agile Deliberation: How AI helps humans define the gray areas
Google Research blog · Nov 5, 03:07 KST
Read on Google Research blog ↗AI summarized the original article. Check the original for accuracy. Copyright belongs to Google Research blog.
More from this field
- Google Brings Duplex Calling to Gemini, Letting AI Phone Businesses for You
Google has moved its Duplex phone-calling technology into the Gemini app, so Gemini can now call businesses for you. The system has changed a lot since 2018, shifting from narrow custom models to a single stack built on Gemini.
Google Research blog · Nov 21, 08:30 KST - Google Releases Project Genie, an AI That Generates Explorable Worlds
Google has released Project Genie, an AI tool that creates interactive worlds you can explore. The video shows how to access it, walks through a generated world and a T-Rex demo, and discusses why this early technology matters.
The AI Advantage · Nov 21, 08:21 KST - Anthropic Employee Says Claude Helped Prove Riemann Hypothesis
An Anthropic employee and mathematician, Levent Alpöge, says he used the company's Claude chatbot to create a detailed proof of the Riemann hypothesis over about six weeks. The proof has not yet been checked, and it is over 250,000 lines long in Lean, a computer language used to confirm maths proofs.
BBC News · Nov 21, 08:08 KST - Cole Medin Live-Streams His AI Software Factory Turning Ideas Into Working Software
Cole Medin runs his AI software factory live, taking rough requests from viewers through planning, building, and checking until working software runs on screen. He explains how it works and what he learned building it. The stream also shows a game the system built.
Cole Medin · Nov 21, 08:08 KST