The "bottleneck" of artificial intelligence refers to the technical problems and bottlenecks in the development of artificial intelligence. Although artificial intelligence technology has made a lot of progress, there are still many technical problems and challenges in the process of realizing universal artificial intelligence, which have brought influence to technological progress.

The "bottleneck" of artificial intelligence is mainly manifested in the following aspects:

1. Data quality problems. Artificial intelligence needs a lot of data for training, but it is a huge challenge to ensure the quality of data not only needs a lot of manpower and material resources, but also needs to solve the problem of data privacy and security.

2. Algorithm problem. At present, the development of artificial intelligence algorithm is still relatively primary. Although some successful algorithms have been developed, there are some shortcomings in some scenarios, such as insufficient accuracy and poor generalization ability.

3. The problem of computing power. Artificial intelligence needs huge computing resources to support model training and reasoning, but under the existing computing power, it is very difficult to realize some complex models, such as convolutional neural networks.

4. credibility. Although artificial intelligence performs well in some scenarios, there are still problems in its interpretability and even credibility, which makes the application of artificial intelligence technology in high-risk fields face challenges.

these problems have obviously affected the technological progress. Many fields have high expectations for artificial intelligence, such as medical care and finance. However, due to the "bottleneck" problem of artificial intelligence, the application of many key scenarios faces great challenges. To solve these "bottlenecks" problems, we need to start from many aspects such as technology, economy and policy. It is also very necessary to continue technical research and innovation, strengthen international cooperation and data security, and establish good policies, laws and regulations. Only by integrating these efforts can we promote the leap-forward development of artificial intelligence technology and make it a powerful tool for human beings.