谷歌旗下的DeepMind公司近日在Nature杂志上发表了一项重大研究成果，他们利用大模型谷歌旗下的DeepMind公司近日在Nature杂志上发表了一项重大研究成果，他们利用大模型解决了困扰数学家60多年的数学难题，这一突破性的解法超出了人类已有的认知。
这项名为FunSearch的技术，其中的“Fun”是函数（Function）一词的简写，利用大模型解决长期存在的科学难题，产生以前不存在的可验证且有价值的新信息。这是第一次有人证明基于大模型的系统可以超越数学家和计算机科学家的认知，这一发现不仅新颖，而且比当今存在的任何其他东西都更有效。
在Nature论文配套的新闻解读中，DeepMind负责人表示：“我们使用大模型的方式是当做创造力引擎”。这一创新的思维方式，为解决复杂的科学问题提供了新的可能性。
DeepMind的这一成果再次证明了人工智能在科学研究中的重要作用。随着技术的不断发展，我们有理由相信，人工智能将在未来的科学研究中发挥更大的作用，帮助我们解决更多的科学难题。
英语如下：
====
“News Headline: DeepMind’s Large Model Cracking 60====
“News Headline: DeepMind’s Large Model Cracking 60-Year-Old Mathematical Problem, Surpassing Human Cognition
Keywords: DeepMind, Mathematical Problem, FunSearch
News Content: Google’s subsidiary company, DeepMind, recently published a major research result in the Nature magazine. They used a large model to solve a mathematical problem that has puzzled mathematicians for over 60 years. This breakthrough solution exceeds human existing cognition.
This technology, named FunSearch, where “Fun” is short for Function, uses large models to solve long-standing scientific problems and generate new verifiable and valuable information that did not exist before. This is the first time someone has demonstrated that a system based on a large model can surpass the cognition of mathematicians and computer scientists. This discovery is not only novel but also more effective than anything else currently available.
In the accompanying news interpretation of the Nature paper, the head of DeepMind said, “We use large models as creativity engines.” This innovative way of thinking provides new possibilities for solving complex scientific problems.
DeepMind’s achievement once again proves the important role of artificial intelligence in scientific research. With the continuous development of technology, we have reason to believe that artificial intelligence will play a greater role in future scientific research and help us solve more scientific problems.”
【来源】https://www.qbitai.com/2023/12/106559.html
Views: 2