Big Brother AI: Large Models Train Smaller Counterparts without Human Assistance

NNicholas December 17, 2023 7:01 AM

A new breakthrough in AI development has allowed large AI models to autonomously create and train smaller models. This advancement could lead to the development of highly efficient, cost-effective small AI tools that could be integrated into everyday devices.

Big AI models generate smaller counterparts

In a significant step forward for AI development, large AI models have been demonstrated to have the capability to autonomously generate and refine smaller AI tools. This breakthrough comes from a collaborative effort between a team of scientists from MIT, several University of California campuses, and AI technology company Aizip. The process, likened to a 'bigger brother' helping its smaller sibling to improve, is seen as the first stage towards self-evolving AI.

Unlike their larger counterparts, which can cost up to $700,000 a day to run, TinyML models are highly cost-effective and portable. These smaller AI models are designed to perform specific tasks and can be integrated into a multitude of devices, transforming everything from facial recognition technology to home appliances. This provides a means of bringing AI technology into everyday life in a practical and affordable manner.

Large-scale AI automates the creation of TinyAI

While the team's focus remains on the development of these small-scale AI systems, they stress that the design and refinement process can now be handled by more intelligent, large-scale AI systems. This could potentially lead to an AI-driven ecosystem in which larger AI models guide the development of smaller, more specialized models.

More articles

Also read

Here are some interesting articles on other sites from our network.