Meta Platforms Unveils Human-Like AI Model for More Accurate Image Analysis: i-JEPA

Meta Platforms Announces ‘Human-Like’ AI Model for Image Completion

Introduction

Meta Platforms, the parent company of Facebook and Instagram, has announced the development of a new ‘human-like’ AI model capable of analyzing and completing images with higher accuracy. This article will delve into the details of this new model and Meta’s open-source research initiatives.

Human-Like AI Model

“Fill in the Missing Parts”

Meta’s new AI model, called i-JEPA, uses prior knowledge of the world to fill in missing parts of images. Unlike other generative AI models, i-JEPA takes a ‘human-like’ approach by considering how human beings would complete a picture and avoids common errors in AI-generated images.

Meta’s Research Initiatives

Research Lab

Meta has become a popular publisher of open-source AI research through its in-house research lab. CEO Mark Zuckerberg believes that sharing models developed by Meta researchers can foster innovation, uncover security vulnerabilities, and reduce costs.

Potential Risks and Drawbacks

Ignore the Warnings

Despite warnings from others in the industry, Meta executives have rejected concerns over the potential risks of this technology.

The Godfather of Artificial Intelligence

Even Jan Lucan, a lead AI scientist at Meta and considered one of the ‘godfathers of artificial intelligence’, refuses to focus on the potential drawbacks of this technology and prioritizes work on security controls.

Integrating AI in Consumer Products

Meta has begun integrating generative AI modeling capabilities into its consumer products, such as advertising tools and an Instagram product that allows users to edit photos based on their searches.

The post Meta Platforms Unveils Human-Like AI Model for More Accurate Image Analysis: i-JEPA appeared first on Asume Tech.

from Technology - Asume Tech https://asumetech.com/meta-platforms-unveils-human-like-ai-model-for-more-accurate-image-analysis-i-jepa/