Unlike existing AI systems, Meta's so-called Image Joint Embedding Predictive Architecture, or I-JEPA, is designed to aid users based on what it already knows about the world. While existing models compare nearby pixels to aid in image creation, Meta's I-JEPA is already fed with an internal model of the outside world which it then uses to determine what it needs to produce. The result? A generative model that is less prone to producing people with too many, too few, or impossibly interconnected fingers.
Meta is heating up the AI arms race with the launch of a new artificial intelligence model that can help users analyze and complete unfinished images. What helps this model stand apart is not necessarily in what it does, but how it does it.
Unlike existing AI systems, Meta's so-called Image Joint Embedding Predictive Architecture, or I-JEPA, is designed to aid users based on what it already knows about the world. While existing models compare nearby pixels to aid in image creation, Meta's I-JEPA is already fed with an internal model of the outside world which it then uses to determine what it needs to produce. The result? A generative model that is less prone to producing people with too many, too few, or impossibly interconnected fingers.
"The idea behind I-JEPA is to predict missing information in an abstract representation that’s more akin to the general understanding people have," the company said in a statement.
And it's all open-source! Meaning, unlike OpenAI's ChatGPT, users will be able to build on top of the models Meta's researchers have developed as part of CEO Mark Zuckerberg's plan to let the community innovate, spot safety gaps, and lower costs.
"For us, it's way better if the industry standardizes on the basic tools that we're using and therefore we can benefit from the improvements that others make," Zuckerberg was quoted telling investors in April.
I-JEPA's features are expected to creep into Meta's other offerings, including Facebook, WhatsApp, and Instagram, so keep your eye peeled for more AI.
Instagram ranked #3 on HackerNoon's Tech Company Rankings this week, followed by Facebook at #4, and Whatsapp on #9.
👋 You’re reading part 1 of HackerNoon's Tech Company News Brief, a weekly collection of tech goodness that combines HackerNoon's proprietary data with internet trends to determine which companies are rising and falling in the public consciousness. Part 2 goes live tomorrow. Hate waiting? No problemo! Just subscribe here to receive the complete newsletter a day early in your inbox.
In Other News.. 📰
Crypto bros, rejoice! A new agreement between Binance and the U.S. Securities & Exchange Commission means customers no longer need to move their assets outside the U.S. until a lawsuit pending in court is resolved.
The U.K. government tapped Ian Hogarth to head the country's AI Foundation Model Taskforce, a working group charged with looking "at the risks surrounding AI." Described as a tech investor, entrepreneur and AI specialist, Hogarth is the co-author of the highly influential annual State of AI report and the founder of the start-up Songkick and the venture capital fund Plural.
Bill Gates is 'friends' with Chinese president Xi Jinping.
AI is helping The Beatles create its final record.
Google has asked its employees to refrain from using chatbots, including its own. Google ranked #55 on HackerNoon's Tech Company Rankings.
A ransomware group has threatened to release 80 gigs of data it stole from Reddit, unless the company pays $4.5 million in ransom and changes its controversial API pricing.
And that's a wrap! Don't forget to share this newsletter with your family and friends!
See y'all next week. PEACE! ☮️
— Sheharyar Khan, Editor, Business Tech @ HackerNoon