Scientists and AI experts have made a significant breakthrough in the field of artificial intelligence: large AI models, like the one powering ChatGPT, are now capable of autonomously creating smaller AI tools. This development marks a new era in AI evolution, where larger models can act as ‘big brothers’ to smaller ones, guiding their development without human intervention.
A collaborative team from MIT, several University of California campuses, and AI technology company Aizip have been at the forefront of this advancement. Yan Sun, CEO of Aizip, explained to Fox News that this is the first step towards self-evolving AI, where AI models can independently build other AI models.
While large language models like ChatGPT are known for their complexity and high operational costs, smaller AI models, often referred to as TinyML, are cost-effective and portable. These tiny AIs can be integrated into various devices and are designed for specific tasks such as facial recognition, hearing aids, and home appliances. They bring intelligence into everyday life, enhancing safety and convenience.
The team’s focus is on creating these tiny AIs, but the process of designing them can now be fully automated by more advanced AI. This signifies a future where large and small AI models collaborate, forming a complete intelligence ecosystem.
Microsoft, in partnership with OpenAI, has also highlighted its small yet powerful AI models that could potentially rival ChatGPT. The ability of large AI models to create and train smaller ones without human help is a groundbreaking step in AI technology, promising a wide range of applications and advancements in the field.
Source: Business Insider