KEY HIGHLIGHTS
- AI Self-Evolution Breakthrough: Scientists in the U.S. have reportedly achieved the first-ever AI capable of autonomously creating smaller AI systems without human input, marking a significant leap in self-evolving AI.
- Larger Models Empowering Smaller Applications: The research reveals that bigger AI models, like those behind ChatGPT, can autonomously generate specialized AI applications. This collaborative approach enhances capabilities for various real-world applications.
- Diverse Applications: Specialized AI models created by this autonomous process can be utilized in diverse fields, from improving hearing aids to monitoring oil pipelines and tracking endangered species, showcasing the versatility of the technology.
- Automated Pipeline Achievement: The scientists have developed a fully automated pipeline for AI model creation, from data generation to deployment and testing, eliminating the need for human intervention in the process.
- Tiny Machine Learning for Compact Devices: The demonstrated human activity tracker, smaller than a dime, exemplifies “tiny machine learning” – compact AI systems crucial for pervasive AI. This technology enables intelligent functionality in small devices and spaces.
- Cloud-based vs. Embedded Models: The breakthrough emphasizes the spectrum of intelligence, with large AI models residing in the cloud (e.g., ChatGPT) and the newly developed small models embedded in devices, paving the way for a comprehensive AI ecosystem.
Breakthrough in Autonomous AI Creation
A team of U.S. scientists claims they’ve made the first-ever artificial intelligence (AI) that can create more AI without needing people to babysit it.
In a report released last Friday, the scientists revealed that AI models can now whip up smaller AI systems all on their own, without any human assistance. This breakthrough project, led by Aizip Inc. and researchers from the Massachusetts Institute of Technology and various University of California campuses, shows that big AI models (like the ones powering ChatGPT) can make specialized AI applications for everyday use. These specific models could be handy for things like improving hearing aids, keeping an eye on oil pipelines, and tracking endangered animals.
Yan Sun, the big boss at Aizip, explained the process, saying, “Right now, we’re using the big models to create the little models, like a big brother helping out a smaller brother to get better. It’s the first step toward these AIs getting smarter on their own. This is the first step in showing that AI models can build AI models.”
Yubei Chen, one of the researchers, agreed with Sun, saying, “The surprising thing we find is that, essentially, you can use the largest model to help you automatically design the smaller ones. So in the future, we believe that these, the large and the small, they will collaborate together and then build a complete intelligence ecosystem.”
AI Models Building Specialized Applications
The spawned AI models can do cool stuff like recognizing human voices in noisy places, keeping an eye on pipelines to stop problems before they happen, and tracking wild animals using data from satellites and ground-based sensors.
Chen highlighted the significance of their technology, stating, “Our tech is a breakthrough in the sense that for the first time, we have designed a fully automated pipeline. It can design an AI model without human intervention in the process.”
He continued, “This month, we just demonstrated the first proof of concept such that one type of model can be automatically designed all the way from data generation to the model deployment and testing without human intervention.”
Sun and Chen showcased a tiny device they made – a human activity tracker that uses AI to watch and understand movement, all packed into a chip smaller than a dime. This tiny machine learning example is crucial for making small AI systems that can fit into little devices or spaces, bringing intelligence to almost anything, from big models like ChatGPT to small gadgets.
Source(s): Fox News
The information above is curated from reliable sources, modified for clarity. Slash Insider is not responsible for its completeness or accuracy. Please refer to the original source for the full article. Views expressed are solely those of the original authors and not necessarily of Slash Insider. We strive to deliver reliable articles but encourage readers to verify details independently.