B
新素材 Sora 视频模型 On Feb. 15, 2024, a new video model called Sora promised to create a video up to a minute long from a short text description. Just tell Sora your idea, and watch it make a video right before your eyes. Sora is described as “impressive” “world - changing” and “breathtaking, yet scary” by people who have watched the video.
The videos from OpenAI's Sora are full of details. One video of a Tokyo street scene shows that Sora has learned how objects fit together in 3 - D: the camera goes into the scene to follow a couple as they walk past a row of shops. Sora knows how things should look in real life.
Sora looks at many videos and learns from them by breaking them into very small bits. It uses these bits to make a new video. The new video is grey at first and very messy. Then, Sora fixes the video until it looks good and smooth. Sora also gets help from ChatGPT, which turns the user's short sentences into clear instructions. This helps Sora make the video just as you asked.
Sora may find “busy” pictures hard to understand. For example, when a person moves, things around him or her change. In one video created by Sora, a grandma blows out the candle on a birthday cake, but the flame(火焰)doesn't move. Sora may not be able to tell “left” from “right”.
OpenAI also said that Sora deals with occlusion well. One problem with the other existing models is that they fail to follow objects when they are out of sight. For example, if a truck passes in front of a street sign, the sign might not reappear afterward.
In a video of a papercraft underwater scene, Sora has added what looks like cuts between different pieces of footage(镜头), and the model has kept the same style between them. But it's not perfect. In the Tokyo video, cars to the left look smaller than the people walking beside them. They also pop(突然出现)in and out between the trees. “There's some work to be done,” said Brooks. “For example, if someone goes out of view for a long time, they won't come back.”
“Sora seems like a very important leap forward,” said Sam Gregory, a human rights advocate that specializes in the use and misuse of video technology. “But there are two sides of the coin,” he said. “The meaningful abilities help many more people to be storytellers using videos. And there are also possibilities for misuse.”
OpenAI has also realized the risks that come with a video model. The OpenAI team plans to draw on the safety testing. “We'll need to get more feedback(反馈)and learn more about the types of risks that need to be solved while using videos before it would make sense for us to put Sora into use,” said Ramesh.
6. What isn't people's attitude towards Sora in Paragraph 1?
A. Amazing. B. Horrible.
C. Important. D. Dangerous.
7. Which of the following words has the closest meaning to “occlusion”?
A. Being seen. B. Being followed.
C. Being covered. D. Being dropped.
8. 新考法补全文章 The following sentence would be best placed at the end of ______.
Perhaps it will be an imagination tool or a misinformation machine.
A. Paragraph 4 B. Paragraph 3
C. Paragraph 2 D. Paragraph 1
9. 思维品质逻辑思维 The next part of the passage may be about ______.
A. the safety testing
B. the misuse of Sora
C. two sides of Sora
D. another video model
10. What can most possibly be the writer's real purpose?
A. To encourage us to use Sora to make short videos.
B. To explain to us how to use Sora to create short videos.
C. To remind us to care about the problems and risks Sora has.
D. To introduce what advantages Sora will bring to our daily life.
新素材 Sora 视频模型 On Feb. 15, 2024, a new video model called Sora promised to create a video up to a minute long from a short text description. Just tell Sora your idea, and watch it make a video right before your eyes. Sora is described as “impressive” “world - changing” and “breathtaking, yet scary” by people who have watched the video.
The videos from OpenAI's Sora are full of details. One video of a Tokyo street scene shows that Sora has learned how objects fit together in 3 - D: the camera goes into the scene to follow a couple as they walk past a row of shops. Sora knows how things should look in real life.
Sora looks at many videos and learns from them by breaking them into very small bits. It uses these bits to make a new video. The new video is grey at first and very messy. Then, Sora fixes the video until it looks good and smooth. Sora also gets help from ChatGPT, which turns the user's short sentences into clear instructions. This helps Sora make the video just as you asked.
Sora may find “busy” pictures hard to understand. For example, when a person moves, things around him or her change. In one video created by Sora, a grandma blows out the candle on a birthday cake, but the flame(火焰)doesn't move. Sora may not be able to tell “left” from “right”.
OpenAI also said that Sora deals with occlusion well. One problem with the other existing models is that they fail to follow objects when they are out of sight. For example, if a truck passes in front of a street sign, the sign might not reappear afterward.
In a video of a papercraft underwater scene, Sora has added what looks like cuts between different pieces of footage(镜头), and the model has kept the same style between them. But it's not perfect. In the Tokyo video, cars to the left look smaller than the people walking beside them. They also pop(突然出现)in and out between the trees. “There's some work to be done,” said Brooks. “For example, if someone goes out of view for a long time, they won't come back.”
“Sora seems like a very important leap forward,” said Sam Gregory, a human rights advocate that specializes in the use and misuse of video technology. “But there are two sides of the coin,” he said. “The meaningful abilities help many more people to be storytellers using videos. And there are also possibilities for misuse.”
OpenAI has also realized the risks that come with a video model. The OpenAI team plans to draw on the safety testing. “We'll need to get more feedback(反馈)and learn more about the types of risks that need to be solved while using videos before it would make sense for us to put Sora into use,” said Ramesh.
6. What isn't people's attitude towards Sora in Paragraph 1?
A. Amazing. B. Horrible.
C. Important. D. Dangerous.
7. Which of the following words has the closest meaning to “occlusion”?
A. Being seen. B. Being followed.
C. Being covered. D. Being dropped.
8. 新考法补全文章 The following sentence would be best placed at the end of ______.
Perhaps it will be an imagination tool or a misinformation machine.
A. Paragraph 4 B. Paragraph 3
C. Paragraph 2 D. Paragraph 1
9. 思维品质逻辑思维 The next part of the passage may be about ______.
A. the safety testing
B. the misuse of Sora
C. two sides of Sora
D. another video model
10. What can most possibly be the writer's real purpose?
A. To encourage us to use Sora to make short videos.
B. To explain to us how to use Sora to create short videos.
C. To remind us to care about the problems and risks Sora has.
D. To introduce what advantages Sora will bring to our daily life.
答案
C
