Most popular now

Meta's Open-Weight Model Brings AI Agents to Personal Computers

Meta model for AI agents on home computers
Інноваційна модель від Meta робить штучний інтелект доступним для особистих комп'ютерів. Photo: НВ — Техно

An Open Model That Lets AI Agents Run on Home Computers

According to НВ — Техно: Photo: Unsplash

On August 10 at 22:40, Meta unveiled Muse Glimmer, a new open-weights model that can run AI agents directly on users' own computers. Boasting 30 billion parameters, Muse Glimmer is the open version of Meta's closed Muse Spark system, which was introduced back in April. The model's weights are released under the Apache 2.0 license.

Key Features of Muse Glimmer

Muse Glimmer runs in an always-on mode, so it can function with or without an internet connection. It supports text and images and has been trained on more than 100 languages. The model is suited for a broad set of tasks, including:

  • invoking tools;
  • writing code;
  • working with files and screenshots.

It can be run on a Mac or PC using just one consumer graphics card.

Broadly distributing superintelligence has the potential to set off a new era of personal empowerment, where people can apply this capability to develop themselves, follow their passions, and improve daily life and the world more than ever before. - Mark Zuckerberg

By releasing Muse Glimmer with open weights, Meta may dramatically change how AI is handled on personal computers. People can now deploy complex models without powerful servers or a constant internet connection, which gives developers, researchers, and hobbyists a simpler path to experimenting with AI. This shift could amplify interest in on-device artificial intelligence tools that prioritize data privacy.

As AI technology continues to evolve, concerns about privacy and data usage are becoming increasingly prevalent. For instance, recent developments indicate that Instagram may utilize your photos without your consent, raising questions about how personal data is leveraged in AI systems. Understanding these implications is crucial as we navigate this new landscape of artificial intelligence.

Read also

Advertisement