Meta and NYU Bet on Raw Video for AI

⚡

Key Takeaways

1Meta and New York University have designed a multimodal AI model, challenging established assumptions.

2Meta researchers are turning to unlabeled video to train their models, due to the scarcity of textual data.

3This strategy could enable the creation of more robust and versatile AI models by leveraging new forms of data.

💡Why it matters — The depletion of textual data is driving tech giants to innovate, influencing the future of AI and its applications.

⚡Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisis ton rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

📄

Full Analysis

Meta and NYU: An Innovative Multimodal AI Model

A collaboration between Meta FAIR and New York University has led to the creation of a multimodal artificial intelligence model built from the ground up. This project has revealed that several preconceived notions about building these models do not hold up against practical application.

Unlabeled Video: A New Resource for AI

As textual data, essential for language models (LLMs), begins to dwindle, Meta is shifting its efforts towards the use of unlabeled videos. This resource is seen as the next big frontier for training AI models. By leveraging this type of data, Meta hopes to develop more robust and versatile models capable of adapting to a variety of tasks.

⚡

Brief IA — L'actualité IA en français

L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.

📰 Voir toutes les actus IA →