share_log

Meta(META.US)推出自我评估AI模型 减少人类参与提升开发效率

Meta (META.US) launches self-assessment AI model to reduce human involvement and improve development efficiency.

Zhitong Finance ·  14:09

Meta, the parent company of Facebook, announced that it will release a series of new artificial intelligence models.

According to the Intelligence Finance APP, Meta (META.US), the parent company of Facebook, announced that it will release a series of new artificial intelligence models, including one that can self-assess, which may reduce human involvement in the artificial intelligence development process. This release comes after Meta introduced the tool in August, detailed in a paper that describes how the tool uses the same 'idea chain' technology as OpenAI's o1 model to improve accuracy in answering challenging questions related to science, coding, and mathematics. Meta's researchers trained the assessment model using data generated entirely by artificial intelligence, eliminating the need for human input.

This move by Meta marks an important step for the company in artificial intelligence self-learning and self-assessment capabilities. This self-assessment technology not only enhances model accuracy but also may reduce costs and inefficiencies in reinforcement learning processes that currently rely on human feedback. Meta's researchers state that assessing its own abilities using artificial intelligence is crucial in creating intelligent agents that can learn from mistakes and improve autonomously. These intelligent agents are envisioned as digital assistants capable of performing numerous tasks without human intervention.

Jason Weston, a researcher at Meta, emphasizes that as artificial intelligence advances, it will become increasingly proficient at self-checking. The ability for self-learning and self-assessment is crucial for artificial intelligence. Apart from Meta, other companies such as Google and Anthropic are also researching reinforcement learning concepts based on artificial intelligence feedback, but they often do not release models publicly.

In this release, Meta also introduced other AI tools, including updates to the image recognition model Segment Anything, which can accelerate response generation times for large language models (LLMs), and a dataset to help discover new inorganic materials. The release of these tools further demonstrates Meta's efforts and commitment to advancing artificial intelligence technology.

The translation is provided by third-party software.


The above content is for informational or educational purposes only and does not constitute any investment advice related to Futu. Although we strive to ensure the truthfulness, accuracy, and originality of all such content, we cannot guarantee it.
    Write a comment