


Matryoshka dolls are not advisable: Researchers confirm that training AI with AI-generated results will lead to model degradation
IT House News on June 14th, IT House friends may have imagined what kind of results can be obtained if the results generated by AI are used to train AI and perform "matryoshka-style training"? There is currently a research team that has observed and recorded this, and detailed papers and results have been published on arXiv.
Summary in one sentence - "Using content generated by the model in training will lead to irreversible defects in the subsequently generated model." In human terms, the researchers found that "training AI with results generated by AI will only make the The models are getting worse."
▲ Picture source arXiv
It is reported that the researchers specifically studied the probability distribution of AI generation models, mainly focusing on "text to text" and "image to image", and finally concluded: "Because the results generated by each model have certain characteristics , so if you train AI with a model generated by AI, over time, the latter will forget the real underlying data distribution."
▲ Picture source arXiv
Ilia Shumailov, one of the main authors of the paper, also said that “over time, errors in the generated data (IT Home Note: such as false examples) will force AI to further misperceive reality. We were surprised to observe that the model Crashes happen quite quickly, and models can quickly forget much of the original data they originally learned from."
But friends may have questions. If the results generated by AI are manually polished and then put into model training, can the model be "degraded"?
The answer is no. The researchers found that "the model degradation process is inevitable", so even for "polished and idealized AI output content", the model will experience certain degradation after long-term learning.
For any large model, due to their excessive learning data, they will inevitably come into contact with data generated by other AIs, so the researchers said that "AI identification should be introduced to pick out learning data that may contain errors." To improve the learning ability and accuracy of the model.
The above is the detailed content of Matryoshka dolls are not advisable: Researchers confirm that training AI with AI-generated results will lead to model degradation. For more information, please follow other related articles on the PHP Chinese website!

Since 2008, I've championed the shared-ride van—initially dubbed the "robotjitney," later the "vansit"—as the future of urban transportation. I foresee these vehicles as the 21st century's next-generation transit solution, surpas

Revolutionizing the Checkout Experience Sam's Club's innovative "Just Go" system builds on its existing AI-powered "Scan & Go" technology, allowing members to scan purchases via the Sam's Club app during their shopping trip.

Nvidia's Enhanced Predictability and New Product Lineup at GTC 2025 Nvidia, a key player in AI infrastructure, is focusing on increased predictability for its clients. This involves consistent product delivery, meeting performance expectations, and

Google's Gemma 2: A Powerful, Efficient Language Model Google's Gemma family of language models, celebrated for efficiency and performance, has expanded with the arrival of Gemma 2. This latest release comprises two models: a 27-billion parameter ver

This Leading with Data episode features Dr. Kirk Borne, a leading data scientist, astrophysicist, and TEDx speaker. A renowned expert in big data, AI, and machine learning, Dr. Borne offers invaluable insights into the current state and future traje

There were some very insightful perspectives in this speech—background information about engineering that showed us why artificial intelligence is so good at supporting people’s physical exercise. I will outline a core idea from each contributor’s perspective to demonstrate three design aspects that are an important part of our exploration of the application of artificial intelligence in sports. Edge devices and raw personal data This idea about artificial intelligence actually contains two components—one related to where we place large language models and the other is related to the differences between our human language and the language that our vital signs “express” when measured in real time. Alexander Amini knows a lot about running and tennis, but he still

Caterpillar's Chief Information Officer and Senior Vice President of IT, Jamie Engstrom, leads a global team of over 2,200 IT professionals across 28 countries. With 26 years at Caterpillar, including four and a half years in her current role, Engst

Google Photos' New Ultra HDR Tool: A Quick Guide Enhance your photos with Google Photos' new Ultra HDR tool, transforming standard images into vibrant, high-dynamic-range masterpieces. Ideal for social media, this tool boosts the impact of any photo,


Hot AI Tools

Undresser.AI Undress
AI-powered app for creating realistic nude photos

AI Clothes Remover
Online AI tool for removing clothes from photos.

Undress AI Tool
Undress images for free

Clothoff.io
AI clothes remover

Video Face Swap
Swap faces in any video effortlessly with our completely free AI face swap tool!

Hot Article

Hot Tools

Notepad++7.3.1
Easy-to-use and free code editor

Dreamweaver Mac version
Visual web development tools

ZendStudio 13.5.1 Mac
Powerful PHP integrated development environment

SAP NetWeaver Server Adapter for Eclipse
Integrate Eclipse with SAP NetWeaver application server.

DVWA
Damn Vulnerable Web App (DVWA) is a PHP/MySQL web application that is very vulnerable. Its main goals are to be an aid for security professionals to test their skills and tools in a legal environment, to help web developers better understand the process of securing web applications, and to help teachers/students teach/learn in a classroom environment Web application security. The goal of DVWA is to practice some of the most common web vulnerabilities through a simple and straightforward interface, with varying degrees of difficulty. Please note that this software